Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

May 2013

1 participants
62 discussions

ffmpeg-devel.log.20130520
by burek 21 May '13

21 May '13

[00:11] <ubitux> i wonder if i couldn't actually just put the v4l2 muxer into v4l2.c [00:11] <ubitux> and avoid the split [00:11] <ubitux> saste: opinion? [00:11] <ubitux> the outdev is pretty trivial [00:12] <ubitux> ~100L of code [00:12] <saste> possible, as you prefer [00:12] <ubitux> it will make a smaller diff [00:14] <ubitux> but might require some ugly ifdefery. [00:17] <ubitux> saste: it's available in my v4l2 branch if you want to retry with skype or w/e [00:30] <saste> ubitux, the problem with skype is that it *opens* the device [00:30] <saste> if the device has been already opened there is no way to access it [00:42] Action: michaelni figured out why haiku fate fails ... corrupted fate reference file ... how sweet [00:44] <ubitux> saste: basically if you stop pushing to the virtual device, you can re-open it for pushing later because skype is blocking the open? [00:44] <ubitux> you can't* [01:08] <ubitux> saste: "push" because of "grab" [01:09] <ubitux> (in the indev) [01:09] <ubitux> should i simply use "output"? [01:18] <ubitux> michaelni: corrupted fate ref file? huh? [01:18] <ubitux> corrupted FS? [01:18] <michaelni> dunno, maybe i copied tge wrong file to it, theres no rsync on haiku [01:19] <michaelni> no fsck either that i could see [01:22] <cone-582> ffmpeg.git 03Clément BSsch 07master:8eec655320f5: lavd/v4l2: move pixel format conversion code to a separate place. [01:22] <cone-582> ffmpeg.git 03Clément BSsch 07master:16a75eaa204a: lavd: add v4l2 outdev. [01:37] <ubitux> someone should fix the transparency in the ffmpeg logo :P [01:37] <ubitux> or maybe it's a problem in our decoder [01:40] <saste> ubitux, there is no bug [01:41] <saste> the correct fix would be to enhance SDL to support alpha images [01:41] <saste> right now we dismiss alpha info when removing alpha, resulting in the corruption [01:41] <ubitux> there is some weird black shapes :p [01:41] <ubitux> btw, [01:41] <ubitux> there is something really strange [01:42] <ubitux> it seems dctdnoiz is not calling the asm optimized version [01:42] <ubitux> that might explain why it's so slow.. [01:42] <cone-582> ffmpeg.git 03Stefano Sabatini 07master:b7be8ea92a4b: lavfi/geq: mark timeline support [01:42] <ubitux> http://ubitux.fr/pub/pics/_dct-dnoiz-cmap.png [01:42] <ubitux> maybe i need to avcodec init something? [01:46] <ubitux> oh or maybe there is just no x86 version? oO [01:47] <ubitux> it seems there is only an ARM version& [01:47] <ubitux> :((( [01:55] <BBB> use libvpx fdcts [01:57] <ubitux> porting effort :( [01:57] <ubitux> but that would be cool yes [02:49] <Compn> ubitux : a v4loutdev means you can play videos as like a webcam input for other v4l programs ? [02:50] <ubitux> it means you can push some video to a v4l2 (often virtual) device [02:51] <ubitux> example: [/dev/video0 or whatever-video] <--(pull)-- ffmpeg --(push)--> virtual /dev/video1 <--(pull)-- skype/ffplay/any webcam app [02:52] <ubitux> (but it seems there is some issue with skype according to saste) [02:52] <ubitux> Compn: now the cool thing is, you can insert some filters in ffmpeg, which you could control with the ØMQ shell [02:53] <ubitux> so you can make some interactive filtering on your video, while in a virtual hangout for instance [02:53] <ubitux> which is cool! [04:38] <cone-582> ffmpeg.git 03Giorgio Vazzana 07master:d0a34aeedff1: md5: optimize second round by using 4-operation form of G() [04:39] <Compn> ubitux : deserves a news page. people ask about that all the time [04:39] <Compn> er a news entry [04:39] <Compn> playing video as an input to internet streams, that is [10:51] <cone-709> ffmpeg.git 03Martin Storsjö 07master:0b6899f15434: configure: msvc: Set the target windows version to XP if no target is set [10:51] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:2afcd16711e0: Merge commit '0b6899f154347cdd0aa66c5ec16ed9c65871766f' [10:57] <cone-709> ffmpeg.git 03Jean-Baptiste Kempf 07master:c360fb174b01: win32: Do not use GetProcAddress when compiled for Vista+ [10:57] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:faac955d9b7a: Merge commit 'c360fb174b01170ab3fd3048d4d0acb44aa4e834' [11:03] <cone-709> ffmpeg.git 03Luca Barbato 07master:98186578a2a2: vorbis: refactor vorbis_residue_decode_internal [11:03] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:fbba76e89d85: Merge commit '98186578a2a21af5bfe6dd71d222dc270f763c7d' [11:13] <cone-709> ffmpeg.git 03Stefano Sabatini 07master:281fc4e95845: doc/filters: extend/fix documentation for the geq filter [11:13] <cone-709> ffmpeg.git 03Stefano Sabatini 07master:6e7c534955d9: doc/filters: add RGB geq example [11:36] <cone-709> ffmpeg.git 03Luca Barbato 07master:36f10380043c: vorbis: use normal integer values for iterators [11:36] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:aef044d215a5: Merge commit '36f10380043c8b8c5cbf8c1add8695089cdaa341' [11:52] <cone-709> ffmpeg.git 03Luca Barbato 07master:ddf1b4a2f8a6: vorbis: simplify the inner loop in setup_classifs [11:52] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:ed113d7a6b2d: Merge commit 'ddf1b4a2f8a680126eb611428e4f47e6e5b8c6c0' [12:07] <cone-709> ffmpeg.git 03Luca Barbato 07master:db347280eb37: vorbis: fallback to normal division instead of crashing [12:07] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:c5e11e897a9d: Merge commit 'db347280eb3746d7b70888db97f1efe5f06f5624' [12:13] <cone-709> ffmpeg.git 03Janne Grunau 07master:4b79668154f3: mpegvideo: unref old current_picture before allocating a new one [12:13] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:d61681f9e99e: Merge remote-tracking branch 'qatar/master' [14:31] <ubitux> michaelni: huh? that function is not inlined by default? [14:31] <ubitux> it's a static used only once :/ [14:31] <ubitux> afaict [14:32] <michaelni> i checked speed & filesize, both chnaged [14:32] <michaelni> i did not look at the asm [14:44] <kwizart> Hi, does cpu-runtime-detection will works with current ffmpeg 1.2.x and arm cpu ?( such as neon detection ?) [14:46] <ubitux> runtime cpu detect is for libpostproc and some stuff in ppc only iirc [14:46] <ubitux> (and there is no arm in libpostproc so far) [14:50] <nevcairiel> doesnt basically all asm use runtime feature selection [14:56] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:9c49d5908d07: avutil/hmac: silence pointer type warnings [16:12] <BBB> nevcairiel: in some antiquated modules it is configurable for some undescribably stupid reason [16:12] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:f099cdaba38d: avcodec/ratecontrol: silence "warning: expected_bits may be used uninitialized in this function" Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [16:33] <cone-709> ffmpeg.git 03Clément BSsch 07master:9a7f1519aac4: lavd/v4l2enc: check write return value. [16:51] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:2187600f47eb: libx264: silence: warning: the address of val will always evaluate as true [21:55] <cone-709> ffmpeg.git 03Michael Niedermayer 07master:2211c76287e0: smacker: remove av_clip_int16() [22:31] <cone-709> ffmpeg.git 03Giorgio Vazzana 07master:a67304d05f11: avutil/md5: move loop inside inner function [00:00] --- Tue May 21 2013

1 0

ffmpeg.log.20130520
by burek 21 May '13

21 May '13

[01:43] <LunaVorax> Hum [01:43] <LunaVorax> I'm unable to for ffmpeg to make a video from multiple png files at 60 fps [01:44] <LunaVorax> It always want to do it at 25 fps [01:44] <LunaVorax> Whenever I use -r 60 or -framerate 60 [01:44] <LunaVorax> I'm using the following command [01:44] <LunaVorax> ffmpeg -f image2 -i %05d.png -c:v libx264 -preset ultrafast -qp 0 test.mkv [01:44] <LunaVorax> (i have intentionnaly removed the framerate parameter) [01:45] <Mavrik> yeah&. that was a mistake :) [01:45] <LunaVorax> Oh great [01:45] <Mavrik> where are you putting the framerate parameter? :) [01:45] <LunaVorax> I figured it out [01:45] <Mavrik> you had to move it in front of the command? ;) [01:45] <LunaVorax> As very first parameter it works [01:45] <Mavrik> yep :) [01:46] <LunaVorax> I have a lot of trouble dealing with the order of the parameters very often with ffmpeg [01:46] <Mavrik> before -i -> input parameters; after -i -> output parameters [01:46] <Mavrik> that's all you need to know more or less [01:46] <Mavrik> and since you've put the framerate parameter in the output section [01:46] <Mavrik> ffmpeg read the input at default fps (25) and then duplicated frames to achieve 60fps :) [01:46] <LunaVorax> I'm breaking my teeth on a DOSBox recording [01:47] <LunaVorax> You have no idea [01:47] <Mavrik> we're video professionals, we have alot of ideas :P [01:47] <LunaVorax> :P [01:47] <LunaVorax> I ended up decomposing the video in PNG frames [01:47] <LunaVorax> Writing a bash script to remove the frame % 7 == 0 [01:47] <LunaVorax> And re assembling it into a video [01:48] <Mavrik> uh [01:48] <LunaVorax> putting the soundtrack and praying it'll work [01:48] <Mavrik> why? [01:48] <LunaVorax> The video track was playing a lot slower than the audio track (playing at normal speed) [01:48] <LunaVorax> I wanted a way to make the video lenght matching the audio lenght [01:49] <LunaVorax> And since tweaking the fps was never working the way it should, I ended up how much frames I should drop for a 60fps movie [01:50] <LunaVorax> My brain is designed to find complex so-so solutions, it's a curse [01:50] <Mavrik> anyway, gotta run :) [01:50] <Mavrik> later :) [01:50] <LunaVorax> Thanks btw [01:50] <Mavrik> g'luck with your records [03:26] <highgod> Hi, I want to ask a quesion, can ffmpeg decoder be set the preset? [03:26] <Mavrik> uh.~ [03:26] <Mavrik> elaborate. [03:27] <highgod> FAST, BALANCED, QUALITY, LOW_LATENCY [03:27] <Mavrik> on decoder? [03:27] <Mavrik> how would that work? :) [03:27] <highgod> yes [03:27] <highgod> So, I ask, I think it is set by encoder [03:28] <highgod> and decoder just decode the stream [03:28] <Mavrik> you are correct :) [03:28] <Mavrik> there is a "flag2 fast" which skips some things in some decoders at the expense of quality [03:28] <Mavrik> but it doesn't really do much in terms of performance gains [03:29] <highgod> and also the Profile, Level, such as h264 [04:48] <Digit_> hi :o) [04:50] <Digit_> i made a java gui program for ffmpeg, at this time we can use it to stream a playlist or vlc to jtv or ustream and i was wondering, when my program will be great if i may ask for some donations or not ? as it uses ffmpeg [09:36] <xlinkz0> i asked this last night but maybe people were sleeping.. does anyone know how to do this with ffmpeg? http://i.imgur.com/ugrtOew.gif [10:02] <Huemac> ERROR: opencl not found [10:02] <Huemac> :( [10:03] <Huemac> damn. no matter what i try the building always fails at that step [10:04] <Huemac> i have an ubuntu build machine and i installed the AMD-APP-SDK which installs to /opt/AMDAPP [10:05] <Huemac> ./configure --extra-cflags=-I/opt/AMDAPP/include --extra-ldflags=-L/opt/AMDAPP/lib/x86 --enable-opencl --enable-cross-compile --arch=x86 --target-os=mingw32 --disable-w32threads [10:26] <easyfab> Huemac: I have the same error with nvidia sdk under msys/mingw, i'm also interested if someone can give correct ./configure parameters . [10:26] <Huemac> okay [10:26] <Huemac> well, for my luck im in no hurry [10:26] <easyfab> And i think my nvidia sdk is ok because i can build x264 with opencl [10:27] <Huemac> we just need ffmpeg for one of our product and i would like to make a separate build machine for it [10:27] <Huemac> just that i could always build the latest revision with minimal effort [10:43] <Keshl> Kay, so something's weird and I can't figure it out. >w> Lemme try to explain the entire story. I'm recording footage from a capture device. Said device dumps its data in .ts files and such. The ultimate goal is to get them encoded in a more bandwidth-friendly solution (typically using "ffmpeg -i foo.bar -vcodec libx264 -preset medium -crf 23 -movflags faststart -pix_fmt yuv420p OUT.mp4"). Now, when I try to convert the .ts files directly, [10:43] <Keshl> things go kaddywonkus. The audio is synced up, but the video runs at what looks to be 24 (or possibly less) FPS. The .ts file its self is 59.94 FPS. Even if I specify -r 59.94 twice, both for the source and destionation, it keeps the slower framerate. I've tried it in multiple video players so I don't think it's a decoder issue. What's weird is, if I have the supplied software that comes with the capture card convert it to .mp4 for me [10:43] <Keshl> first, everything works just fine. The overall quality is lowered though since it just zips through it and introduces noise, so I'd like to avoid that step if possible. Can anyone tell me why ffmpeg's producing such odd output? [10:46] <Keshl> Oh, forgot to mention. You can poke the file in question (and the outputted .mp4 I get) from ftp://broken@connect.archgeekit.com:21918 -- Just use any password if it asks. [10:58] <Huemac> got it to work [10:59] <Huemac> changed --extra-ldflags=-L/opt/AMDAPP/lib/x86 to --extra-ldflags=-L/opt/AMDAPP/lib/x86_64 :) [11:50] <Hans_Henrik> can i do like "IF src audio codec is of type mp3/aac, copy the audio, if not, convert to aac;" "IF src video codec is x264, copy the video, if not, convert to x264" ? [11:52] <Hans_Henrik> dam g2g srry, ill be back l8r tho [11:52] <JEEB> you'd have to check first with ffprobe or so [11:52] <JEEB> and then handle that in your own script [12:57] <microchip_> how to use the native telecine filter in ffmpeg? [13:00] <ubitux> -vf telecine [13:01] <ubitux> https://ffmpeg.org/ffmpeg-filters.html#telecine [13:10] <microchip_> thanks [15:02] <memento> Hi guys! [15:05] <memento> when converting stream with ffmpeg and passing it to a pipe, I got so many lost frames when playing that pipe with ffplay [15:05] <memento> my commands are (for converting): ffmpeg -i "rtsp://192.168.0.77/channel1" -s 320x240 -f mpegts -y \ff5 [15:06] <memento> for playing: ffplay -f mpegts \ff5 [15:48] <sybariten> memento: have to admit, ive never thought of doing those things with stdin/stdout [15:53] Action: kaliosas b.uy cs, nfs, cod, dirt, nba and other steam acc/gifts. /q me [16:06] <memento> @sybariten, thanks for reply, what I'm trying to do is, to produce thumbnails of live streams, we have 40+ live streams coming to our server with some of them having high quality, to show all or a large portion of them at same time in a grid manner, we must resize them, but doing so, will not reduce system usage, hence we have a poor quality in our software [16:09] <memento> so, I first thought re-streaming with reduced resolution and quality will be enough, but re-streaming will introduce some delay and using vlc to achieve that, will consume more system resource than just showing them. [18:08] <sybariten> memento: sorry, i wasnt here to read your reply.... hope you get the issue sorted out [18:17] <Hans_Henrik> can i do like "IF src audio codec is of type mp3/aac, copy the audio, if not, convert to aac;" "IF src video codec is x264, copy the video, if not, convert to x264" ? [18:18] <xlinkz0> yes, make a script [18:19] <Hans_Henrik> and... how can i make ffmpeg just give me the codec details and do nothing? [18:32] <xlinkz0> Hans_Henrik: http://ffmpeg.org/ffprobe.html [18:33] <JEEB> and yes Hans_Henrik -- I commented on that before you left :P <JEEB> you'd have to check first with ffprobe or so [18:33] <JEEB> <JEEB> and then handle that in your own script [18:35] <Hans_Henrik> JEEB, i didn't get that; [18:35] <Hans_Henrik> <Hans_Henrik> dam g2g srry, ill be back l8r tho [18:35] <Hans_Henrik> * Disconnected (No such device or address) [18:36] <xlinkz0> can you do this with ffmpeg? http://en.wikipedia.org/wiki/Wipe_%28transition%29 [18:40] <Emmanuel_Chanel> Hello! [18:45] <daedeloth> greetings! I want to make a video of 30 seconds, where at second 0, the video is completely blurry and at second 30, the video is crystal clear. Would something liek this be possible with ffmpeg? [19:09] <teratorn> daedeloth: uhm sure anything is possible :) [19:18] <daedeloth> teratorn, yea, but is it possible with ffmpeg? :p [19:22] <mystica555_> ok so im really not certain what the hell im doing wrong... i simply want to convert a 25fps video - Originally NTSC-FILM then timecompressed/audio modified and up-pitched to PAL back to 24000/1001 fps [19:23] <mystica555_> ive gone and modified the audio with audacity which i believe uses a better resampling library for such - made it 0.96 speed [19:23] <mystica555_> i then extracted the h264 with mkvextract to a h264 elementary file [19:24] <mystica555_> im attempting to 'ffmpeg -r 24000/1001 -i file.h264' and adding -r 24000/1001 file.out.mkv at the end too [19:24] <mystica555_> why is mplayer2 still thinking my output is 25fps video??! [19:25] <mystica555_> hell for that matter why is ffmpeg seeing it as 25fps when : the bloody h264 file -seemingly- has no timebase info; mplayer2 on it by itself just decodes as fast as my cpu handles [19:25] <mystica555_> Stream #0:0: Video: h264 (High), yuv420p, 1920x1080, 25 fps, 25 tbr, 1200k tbn, 50 tbc [19:25] <mystica555_> i mean i have every video frame [19:26] <mystica555_> i just want them to play 0.96 slower and match the 0.96 slower audio [19:26] <mystica555_> wtf [19:26] <mystica555_> why is it demanding that it is 25fps [19:26] <teratorn> daedeloth: well you would need a blur filter that changes its blur setting each frame [19:27] <teratorn> daedeloth: I do not know if such a thing exists out of the box (I doubt it) [19:33] <xlinkz0> is it normal to have concat errors when trying to concat h264 files that were created with different presets? [19:34] <xlinkz0> i took the same video, transcoded it with -c:v libx264 -preset slow/fast/superfast , tried to concat them using the concat demuxer and i get this http://codepad.org/YYjI2UPu [19:47] <bunniefoofoo> I am using libavformat to concatenate and remux two mpeg2 streams, but when I play back the streams there is a pause at the seam. The 2nd stream starts on an Iframe and has the same encoding settings as the other (VBR MPEG2 24Mb 29.97 fps) [19:48] <mzso> Hello! [19:48] <mzso> Anyone have an idea how to get ffmpeg/libx264 to get to encode a lossless FPS1 source while keeping tha 0-255 PC range intact? [19:48] <bunniefoofoo> This was not aproblem until I tried using higher bitrates and resolutions [19:50] <Anaphaxeton> any idea about how to create a video stream from a dvdsub? [19:50] <bunniefoofoo> ok it turns out the 1st clip did not have bframes enabled and the 2nd one did, when I corrected that the pause corrected itself, any ideas? [20:44] <tlhiv_laptop> i have a sequence of JPG's that I have converted to a video, and I would like to take another video that I have and place it in the upper right corner of my "slideshow" video [20:45] <tlhiv_laptop> assuming that the JPG slideshow video is 1200x900 and I have another video (4:3 aspect ratio) that i would like to resize to 240:180 and place it in the upper right corner of the 1200:900 [21:17] <brontosaurusrex> any "best" ways to go from 50p to 25p? [21:18] <brontosaurusrex> select_even type of solutions are out of question [00:00] --- Tue May 21 2013

1 0

ffmpeg-devel.log.20130519
by burek 20 May '13

20 May '13

[02:00] <cone-292> ffmpeg.git 03Carl Eugen Hoyos 07master:e32bbd411242: proresdec2: Only try to decode alpha plane if four output planes were allocated. [09:20] <nevcairiel> ftp support broke fate on msvc, btw [10:04] <cone-559> ffmpeg.git 03Hendrik Leppkes 07master:df07a833f1d1: ftp: favor strtoll over atoll [10:20] <cone-559> ffmpeg.git 03Luca Barbato 07master:0f689a3d97a1: wavpack: K&R formatting cosmetics [10:20] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:02788d2ac5df: Merge commit '0f689a3d97a1d3053ca0e9516b296909a72ff31e' [10:26] <cone-559> ffmpeg.git 03Luca Barbato 07master:8c34558131d8: wavpack: return meaningful errors [10:26] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:2ad415cfa420: Merge commit '8c34558131d846d2b10389564caadaa206372fd4' [10:46] <cone-559> ffmpeg.git 03Luca Barbato 07master:ed5067306695: wavpack: validate samples size parsed in wavpack_decode_block [10:46] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:3e3514bce269: Merge commit 'ed50673066956d6f2201a57c3254569f2ab08d9d' [11:06] <cone-559> ffmpeg.git 03Luca Barbato 07master:edfe609fc946: configure: Add basic valgrind-massif support [11:06] <cone-559> ffmpeg.git 03Kostya Shishkov 07master:8e673efc6f5b: prores: update FATE test to account for alpha plane present in the test sample [11:06] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:fa6001e7281d: Merge commit '8e673efc6f5b7a095557664660305148f2788d30' [11:12] <cone-559> ffmpeg.git 03Kostya Shishkov 07master:e45d6222cc92: Escape 130 FATE test [11:12] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:0a8e86faf24f: Merge remote-tracking branch 'qatar/master' [11:44] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:73920ac57340: ffserver: replace atoll() by strtoll() [11:44] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:f51263e6fe29: configure: fix valgrind-massif to actually work [11:44] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:1de28cb7e49b: configure: use the user specified valgrind excutable for massif [11:45] <ubitux> erm [11:46] <ubitux> i think there is a huge memleak, either in nut or in ffv1 [12:12] <nevcairiel> valgrind seems to be happy on fate at least [12:13] <ubitux> maybe it's freed at the end [12:13] <nevcairiel> not a leak then! :) [12:13] <ubitux> well, a membomb if you prefer then [12:14] <ubitux> i membombed myself two times in a row a few minutes ago [12:35] <Compn> ubitux : oom ? [12:35] <ubitux> looks like it's ffv1 related [12:35] <ubitux> Compn: it never works [12:35] <Compn> i mean did you hit out of memory error ? [12:37] <ubitux> yes, memory filled at about 50MB/sec, so my PC freezed quickly [12:48] <ubitux> it's a regression btw [13:07] <ubitux> looks like a regression since a0c0900 [13:07] <ubitux> michaelni ^ [14:14] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:a3264b4ba1e5: ffv1dec: defuse membomb [14:15] <ubitux> haha thx michaelni ;) [14:59] <ubitux> saste: just for you http://prog21.dadgum.com/172.html ;) [15:16] <saste> ubitux: i'm waiting for your zmqshell script ;-) [15:17] <ubitux> oh right, i need to do that.. [15:17] <saste> but talking about escaping, there is an issue in case the command contains a newline [15:17] <ubitux> i'll do it maybe in a few hours [16:09] <cone-559> ffmpeg.git 03Paul B Mahol 07master:f4b2d3598b47: swscale/output: remove unused header [16:09] <cone-559> ffmpeg.git 03Paul B Mahol 07master:b154d203b1c2: swscale/output: silence warning about undefined 'ASSERT_LEVEL' [16:37] <cone-559> ffmpeg.git 03Paul B Mahol 07master:68c4633fa88b: lavfi/histogram: cache log2 of max_hval [17:57] <Compn> ehe [17:58] <Compn> defuse membomb :D [17:59] <cone-559> ffmpeg.git 03Clément BSsch 07master:56f7e29dba39: tools: add zmqshell.py. [18:00] <ubitux> saste: here you go ^ [18:12] <cone-559> ffmpeg.git 03Piotr Bandurski 07master:2f75dcdc1c89: mjpegdec: Fix field orderforinterlacedAVDJand AVRN in ntsc mode [18:12] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:73b9f31c74cc: mjpeg_decode_com: remove redundant len checks [18:45] <ubitux> saste: what's your opinion on #2559? [19:39] <cone-559> ffmpeg.git 03Stefano Sabatini 07master:90cb081f88e2: build: drop unused ALLMANPAGES variable [19:39] <cone-559> ffmpeg.git 03Stefano Sabatini 07master:be3f4c13b6bb: doc/Makefile: install .3 man pages to man3 subdir [19:39] <cone-559> ffmpeg.git 03Stefano Sabatini 07master:66f32a8a7f78: lavfi/drawgrid: fix drawgrid options after 4c205f42c86 [19:48] <saste> ubitux: http://pastie.org/7930203 [19:52] <saste> supposedly it is meant to be a sort of demo [19:55] <ubitux> i get a No such filter: '' [19:55] <saste> btw any reason why geq does not support timeline? [19:56] <ubitux> eval i guess [19:56] <ubitux> but i think it doesn't make much sense [19:56] <ubitux> and we can add timeline indeed [19:57] <saste> ubitux: what does not make sense? [19:57] <ubitux> to prevent timeline with filters with eval [19:57] <saste> http://pastie.org/7930233 [19:57] <ubitux> at first i was concerned about potentially state kept into the eval context [19:58] <ubitux> but for random it doesn't make sense, frame counter don't need to be incremented, and time is absolute everytime as well, so yeah we could add timeline [19:59] <ubitux> hehe that's nice [19:59] <ubitux> now blend an animated banana! [20:00] <saste> i must add an ffmpeg logo [20:00] <saste> i could enable/disable the blend as well [20:00] <ubitux> you should add a v4l2 muxer too [20:00] <saste> ?? [20:00] <ubitux> so we can push to virtual webcam devices with ffmpeg [20:00] <ubitux> saste: https://github.com/umlaeute/v4l2loopback [20:01] <ubitux> saste: https://ffmpeg.org/trac/ffmpeg/ticket/2370 i had to use gstreamer here :( [20:01] <ubitux> even though i wanted to push our life [20:01] <ubitux> gst-launch-0.10 videotestsrc ! v4l2sink device=/dev/video0 [20:02] <saste> ubitux: i implemented that device a long time ago [20:02] <ubitux> :o [20:02] <saste> then i lost it, as i was not very satisfied for some reason [20:03] <saste> if you want to work on it i'll search it on my hd [20:03] <ubitux> no i'd prefer if you get done with it ;) [20:03] <saste> no i'm no good at finishing stuff [20:04] <ubitux> :( [20:09] <saste> ubitux, http://pastie.org/7930298 [20:11] <saste> http://pastie.org/7930311 [20:12] <saste> there is a preparatory patch which does not apply anymore [20:12] <ubitux> and i guess you have no idea what was lacking? [20:13] <saste> ubitux, no [20:13] <saste> it was not always working [20:13] <ubitux> :( [20:13] <saste> in particular i could not make it work with skype, which was my initial objective [20:14] <ubitux> arg and it's gpl [20:15] <ubitux> i guess i'll do it based on something else [20:17] <saste> ah and I didn't know how to select the device [20:17] <saste> right now there is a hardcoded /dev/video1 [20:19] <ubitux> i'm going to try to write one on my own [20:22] <saste> ubitux, I had also this patch: http://pastie.org/7930344 [20:22] <saste> but a lot of things changed since then [20:26] <ubitux> ok :) [20:26] <ubitux> thx [20:27] <cone-559> ffmpeg.git 03Stefano Sabatini 07master:427b51269e15: cmdutils: show timeline support in -filters listing [20:39] <cone-559> ffmpeg.git 03Michael Niedermayer 07master:1a5f1bcff0a6: configure: add more allocation functions to valgrind-massif [20:41] <ubitux> michaelni: do you want me to add a massif instance? [20:41] <ubitux> (does it make any difference? [20:41] <ubitux> ) [20:41] Action: ubitux have no idea what massif is [20:42] <ubitux> maybe it makes no sense actually [20:43] <michaelni> massif generates lots of files that are eaten by an animal called massif-visualizer for example [20:44] <michaelni> and that generates pretty graphs of heap allocations [20:44] <ubitux> mmh [20:44] <ubitux> so it can help defusing more bombs [20:44] <michaelni> yes [21:57] <ubitux> ok, got the v4l2 "muxer" working [21:57] <ubitux> except real time [21:57] <ubitux> oh well, seems to work actually [22:00] <JEEB> 'grats [22:09] <ubitux> timing are broken :p [22:24] <saste> "geq=r=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):g=128 + 78*sin(2*(PI/100)*(cos(PI/3)*(X-32*T) + sin(PI/2)*Y)):b=128 + 132*sin(2*(PI/3)*(cos(PI/12)*(X-50*T) + sin(PI/7)*Y))" [22:24] <saste> endless beautiful variants are possible [22:30] <xlinkz0> can i use ffmpeg with the frei0r mixers? [22:53] <ubitux> saste: that will be really nice with the v4l2 muxer :) [22:54] <ubitux> we will be able to apply filter on our webcam with any app, on the fly with the zmq shell :)) [22:57] <ubitux> saste: oh nice one with geq [22:57] <ubitux> (the color thing) [23:12] <ubitux> there is a regression in the v4l2 indev [23:12] <ubitux> or maybe in ffplay [23:12] <ubitux> the timing are broken [23:12] <ubitux> i guess i'll have to bisect... [23:32] <ubitux> ah, that's libv4l2... [23:44] <cone-582> ffmpeg.git 03Michael Niedermayer 07master:729fa55b8f42: avformat/matroskadec: fix integer overflow [00:00] --- Mon May 20 2013

1 0

ffmpeg.log.20130519
by burek 20 May '13

20 May '13

[00:49] <Anaphaxeton> hello [00:50] <Anaphaxeton> i generated i am trying to extract and add subs to a .mkv target [00:50] <Anaphaxeton> all is fine, syncing and everything [00:50] <Anaphaxeton> except one thing [00:50] <Anaphaxeton> the subs are yellow! [00:50] <Anaphaxeton> how could i correct that? [00:54] <Digit_> ? [00:54] <Digit_> why it says i left [00:56] <Anaphaxeton> (Ping timeout: 250 seconds) [00:58] <Digit_> :( [10:01] <Anaphaxeton> how could i correct the yellow subs i get when muxing the subs from an mkv file? [10:14] <Anaphaxeton> these are dvdsubs. examining the vlc playback of the now embedded (and yellow :p) subs, i see there are only two colors [10:14] <Anaphaxeton> yellow and back [10:14] <Anaphaxeton> i am looking at filter_graphs for chaning those specific colours [10:14] <Anaphaxeton> changing* [10:16] <Anaphaxeton> and since the subs suck from the non-aliased shape i would like to do something about that too [10:31] <ubitux> i'm not sure you can do that from the cli [10:31] <ubitux> api wise, you have access to the palette in the extradata, so you can edit that [10:32] <ubitux> subtitles are not injected in libavfilter, so you can't really do that; though we have a hack for bitmap subs, so maybe you could apply a filter [10:32] <Anaphaxeton> since they are solid colours i thought about just playing with hue. as for anti-aliasing the only thing i saw that could do something is blur...... [10:33] <Anaphaxeton> now i am trying to grab a frame from vlc to determine the hue angle [10:33] <Anaphaxeton> but it creates faulty snapshots... [11:13] <Anaphaxeton> it seems i need to change the rgb value for a specific colour [11:13] <Anaphaxeton> and maybe add a bit of alpha in the subs black [11:13] <Anaphaxeton> am i asking too much? :p [11:21] <Anaphaxeton> lumrgb seems to be the trick but how do i apply it for a specific color? the sub stream has alpha for the whole picture and two colours. yellow borders and black content for the subs [11:21] <Anaphaxeton> i dont want to spoil black by changing yellow [11:51] <esing> Hi, what input and output fps should I set when 'video grabbing' an interlaced PAL vhs tape with preserving the interlacing? [12:08] <xlinkz0> esing: https://en.wikipedia.org/wiki/Video#Number_of_frames_per_second [14:55] <esing> thanks for the link [14:56] <esing> How do I crop in ffmpeg the bottom of a video, which I want to have in 720x576 resolution? I tried crop=720:557:0:20 but that crops the top of my video and the green control line artifact (at the bottom) of my video remains [14:59] <Magicking> esing: Have you tried the documentation of ffmpeg ? [15:01] <esing> Yes, but I don't understand the explanation with crop [=b:h:x:y] b for the cut width , h for the cut height and x,y for the position of the image, default in the middle [15:09] <esing> http://paste2.org/GtkAJ59J [15:14] <doub> esing: the x and y are the difference between the top left of the output relatively to the top left of the input [15:14] <doub> so if you want your output to be the top part of the input, use 0 as y [15:44] <xlinkz0> is there some utility for frei0r that can list all the plugins available? [15:46] <Digit_> halloah :) [15:47] <esing> thanks doub [15:59] <xlinkz0> there's this example in the documentation for the 'geq' filter , how do i test it? nullsrc=s=256x256,geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128 [16:00] <ubitux> ffplay -f lavfi nullsrc=s=256x256 -vf 'geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128' [16:02] <Digit_> not sure i understand the question... [16:02] <xlinkz0> ubitux: i get No such filter: 'geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)2*1000000*sin(N*0.02):128:128' [16:02] <ubitux> then upgrade [16:02] <ubitux> and make sure you have a --enable-gpl build [16:02] <xlinkz0> ah it's windows, it only takes " [16:03] <ubitux> ah misread your paste [16:11] <Digit_> i use an overlay filter with a gif image, is it possible to overlay the image only 2 minutes every 20 minutes ? [16:11] <Digit_> is that clear bcs my english sucks ? :D [16:11] <saste> Digit_, yes, grep for "timeline" in the filters manual [16:12] <saste> btw you need a fairly recent version [16:12] <Digit_> ohhh it's possible ! great ! :) thanks :o) [16:13] <saste> Digit_, overlay=enable=<complicate expression> [16:14] <Digit_> houla i don't know that syntax i am using a filter complex, i have to learn more :) thanks [16:15] <xlinkz0> ubitux: could you help me with this one aswell please? frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [16:15] <Digit_> but i am proud of me i think my gui program starts to be great :) [16:15] <xlinkz0> tried - ffplay -f lavfi frei0r_src=size=200x200:framerate=10 -vf "frei0r=filter_name=partik0l:filter_params=1234" [16:17] <Digit_> what's the prob with that ? [16:18] <xlinkz0> http://codepad.org/nWYCP0vB [16:22] <Digit_> did you try with " " after lavfi ? [16:22] <Digit_> i mean -f [16:22] <Digit_> " after -f and " before -vf [16:26] <Digit_> gulp no after lavfi [16:31] <Digit_> oh but it says the size is invalid for frei0r_src [16:33] <ubitux> xlinkz0: -f lavfi testsrc -vf frei0r=... [16:33] <ubitux> ah you use frei0r_src [16:33] <ubitux> dunno [16:33] <ubitux> should work at first glance [16:34] <ubitux> i don't know frei0r enough [16:35] <esing> Do you recommend to scale a cropped video (720x567) to 720x576 for burning it on dvd? [16:37] <klaxa> letterboxes maybe? [16:37] <klaxa> aspect ratio, etc. [16:42] <lkiesow> Hi, does anone know if it is possible with ffmpeg to extract an image from a video stream every x seconds and overwrite the old files? [16:42] <lkiesow> I know that I can do something like: ffmpeg ... -i /dev/video0 -f image2 -r 1 out%d.png [16:42] <lkiesow> But this will create a lot of files. I only want one image file which is updated regularly. [16:42] <ubitux> -update option [16:43] <ubitux> -f image2 is not required btw [16:44] <lkiesow> ah, thanks a lot & I'll try that [16:45] <lkiesow> since when is there a -update option? My ffmpeg complains about that [16:46] <ubitux> relatively recent [16:46] <ubitux> try -updatefirst maybe [16:47] <ubitux> (exactly the same, the -update alias was added recently) [16:47] <lkiesow> Unrecognized option 'updatefirst'. [16:47] <ubitux> where did you put it? [16:47] <ubitux> what version are you using? [16:51] <ubitux> the option was added in november 2011 [16:51] <ubitux> so you're likely using a very old version [16:51] <ubitux> misplacing the option [16:51] <ubitux> or using a fork [16:53] <saste> ubitux: how can I see which filters support timeline? [16:53] <saste> i mean without grepping the code [16:53] <ubitux> you can tell if the filter supports timeline with the help [16:53] <ubitux> -h filter=curves [16:53] <lkiesow> ubitux: Yes, you were right. I missplaced it. It works. Thanks again :) [16:53] <ubitux> but you can't list them afaik [16:54] <saste> uhm ok [16:55] <ubitux> saste: it might be relevant to add a flag or something in -filters [16:55] <saste> yes [17:08] <esing> klaxa, I think it is even an advantage that I scale (stretch the) video horizontal, because I think that the source is actually 4:3 and I specified 16:9 because I wanted it on my lcd in fullscreen [17:21] <saste> ubitux: should I introduce "FLAGS" like in in -encoders? [17:22] <ubitux> saste: possibly yes [17:28] <saste> ubitux: http://paste.org/64680 [17:29] <ubitux> i would add a "Flags" list below the "Filters" thing but why not [17:29] <ubitux> looks sane so far [17:30] <ubitux> (btw, please don't use paste.org, that's one of the worse you could pick) [17:38] <Anaphaxeton> this works $ ffmpeg -i ~/downton/dt7subs.idx -i ~/downton/dt7wsubs.mkv -map 0:s -map 1:a -map 1:v -c:v h264 -b:v 2500K -c:a copy -c:s dvdsub -filter_complex "[1:v][0:s]overlay[v]" -map [v] withsubs.mkv [17:39] <Anaphaxeton> but the subs are dim [17:39] <Anaphaxeton> i wanted to add a filter to make it more bright [17:40] <Anaphaxeton> i have tried to use something in this fashion: "[s]lutyuv='y=2*val'" [17:40] <Anaphaxeton> 0:s or mapping i have tried the best i have gotten is a brighter movie (the whole picture) [17:41] <Anaphaxeton> whatever i have tried* [17:47] <ubitux> Anaphaxeton: you're doing two things here [17:47] <Anaphaxeton> yes [17:47] <ubitux> muxing the subtitles and hard burning it [17:48] <ubitux> so you likely have a file where the two subtitles overlap themselves [17:48] <ubitux> if you want to change the colors, that's simple [17:48] <ubitux> edit the .idx [17:48] <ubitux> and change the palette [17:48] <Anaphaxeton> one moment [17:48] <Anaphaxeton> you said overlap [17:49] <ubitux> yes, you burned the subtitles onto the video [17:49] <Anaphaxeton> and you made me notice the mkv input [17:49] <ubitux> and you also added it at the format level [17:49] <ubitux> well and yeah you might also have some other subtitles in the input, no idea [17:53] <Anaphaxeton> i am re-muxing the video+audio [17:53] <Anaphaxeton> done [17:53] <Anaphaxeton> i ve never regreted these SSDs! [17:56] <Anaphaxeton> let us see [17:56] <Anaphaxeton> it makes sense i guess if ffmpeg double embeds the subs [17:56] <Anaphaxeton> the dim subs i mean [17:58] <Anaphaxeton> grandiose! now they are yellow... [17:58] <Anaphaxeton> and i havent touched the idx for a while [17:59] <Anaphaxeton> the original # palette: 000000, 0000ff, 00ff00, ff0000, ffff00, ff00ff, 00ffff, ffffff, 808000, 8080ff, 800080, 80ff80, 008080, ff8080, 555555, aaaaaa [18:00] <Anaphaxeton> another one # palette: 000000, 828282, 828282, 828282, 828282, 828282, 828282, ffffff, 828282, bababa, 828282, 828282, 828282, 828282, 828282, 828282 [18:00] <Anaphaxeton> tbh i am clueless about those palettes [18:00] <Anaphaxeton> i can find a reference [18:00] <Anaphaxeton> cant [18:01] <ubitux> dvd subs are using 4-bit colors, and you define those colors through this palette [18:01] <ubitux> that's basic rgb hexadecimal colors [18:04] <Anaphaxeton> and to what does each one correspond? i think this dvd has only two colours. outline and filling [18:04] <Anaphaxeton> and alpha [18:05] <Anaphaxeton> it is badly made... so i need to fix it for the tv to play it [18:06] <Anaphaxeton> imagine that i had to cat >> from vts 0 to vts 2 to be able to extract the subs [18:07] <Anaphaxeton> vlc shows them fine, mplayer shows them black and ffmpeg embeds them as yellow... [18:09] <Anaphaxeton> i am tempted to make a video stream out of the subs, then change hue and luminance and then mux this video with the real one [18:31] <Digit_> that's a lot of things o_O [18:31] <Digit_> first use map then a filter_complex :) [19:16] <Digit_> i mage a gui that uses two ffmpeg processes one is sending to 127.0.01 , the second one with the first as source, to jtv with rtmp, i have to wait many seconds before i launch the second one or i get no image and sometimes i wait fthe same amount of seconds and i get no image aigain , what could be the problem what should i check ? [19:57] <bigmac> im building a batch script, my ffmpeg command shows me a error [19:57] <bigmac> [matroska,webm @ 0xa0c9a0] Read error544kB time=229.48 bitrate= 197.9kbits/s [19:58] <bigmac> http://pastebin.com/uiuAtCSb [19:59] <TheSchaf> :< [19:59] <bigmac> could this be a hard drive problem? [19:59] <saste> TheSchaf, i was faster :) [20:00] Action: TheSchaf cries [20:02] <bigmac> i still get a read error [20:02] <bigmac> avconv [20:03] <TheSchaf> if you have problems with avconv complain in their irc :D [20:03] <TheSchaf> this is ffmpeg [20:04] <sacarasc> bigmac: #libav [20:04] <bigmac> lol [20:04] <bigmac> i just read the thread you gave me [20:04] <bigmac> im not sure what my problem is [20:05] <TheSchaf> but you are using the ffmpeg fork from libav [20:05] <TheSchaf> so you are not using the program this channel is for [20:05] <bigmac> wierd [20:05] <bigmac> lol [20:05] <bigmac> i guess i have to compile from source? [20:05] <JEEB> the error is still clearly visible :P [20:05] <JEEB> "The bitrate parameter is set too low.It takes bits/s as argument, not kbits/s" [20:06] <ubitux> bigmac: ^ [20:06] <ubitux> you can try this first [20:06] <JEEB> although the read error does sound funky [20:06] <JEEB> but first make the low bit rate message go away :P [20:08] <JEEB> both libav's avconv and ffmpeg's ffmpeg support putting a K or M after the number to mark kilo or mega [21:09] <funyun> hi. can ffmpeg be used to add an image to a movie? like when i add my videos to itunes, it just shows a snapshot of some part of the video. but can i set these videos to show an image with ffmpeg? [22:00] <xlinkz0> funyun: if you're asking if you can add a frame in the middle of a movie then yes you can do it [22:00] <xlinkz0> idk how itunes selects the frame for the snapshot [22:00] <xlinkz0> might want to find that out first [22:42] <xlinkz0> can't use frei0r.. [22:45] <xlinkz0> http://i.imgur.com/ewGHv7i.png [22:46] <tyn> If I have a stero audio file and I use the option -ac 1, reducing it to mono, are the channels mixed together into one or is one of the discarded? [22:49] <JEEB> mixed [22:51] <tyn> Thanks a lot, JEEB. [22:56] <bigmac> http://ffmpeginstaller.com/ [22:56] <bigmac> has any one used this [23:05] <JEEB> ahaha [23:06] <JEEB> I haven't looked at the script, but I bet from the looks of the main site that it's rather out-of-date :) [23:06] <JEEB> listing a52 n' such [23:08] <LunaVorax> Hello [23:08] <LunaVorax> I'm having a slight trouble with a video file I have recorded [23:08] <LunaVorax> The video plays slower than the audio track that plays normally [23:09] <LunaVorax> Is there an option in ffmpeg to work the file and as the same time tells that the video should play at such a speed that it isn't longer than the audio track? [23:16] <xlinkz0> how can i do this with ffmpeg? http://www.edmondweblog.com/wp-content/uploads/nito.gif [23:17] <ubitux> 403 [23:17] <xlinkz0> ubitux: http://i.imgur.com/ugrtOew.gif [23:32] <LunaVorax> Hum [23:32] <LunaVorax> Let's put this another way [23:43] <LunaVorax> Does FFMpeg have a feature to drop frames in the video track so it's lenght matches the length of the audio track? [00:00] --- Mon May 20 2013

1 0

ffmpeg-devel.log.20130518
by burek 19 May '13

19 May '13

[01:40] <Compn> lol @ warzones & drones. what have i done?! [01:42] <iive> Compn: ? [01:42] <Compn> iive : Re: [FFmpeg-devel] Patch for Trac ticket #2579 [01:45] <iive> oh... [01:45] <Compn> then read ticket 2579 [01:48] <Compn> michaelni : does ffplay need something similar to "ffmpeg: free threads on error conditions." ? [02:02] <Compn> or anyone else who feels like looking at that... [02:02] <Compn> just curious :) [02:05] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:5918b7ac4116: sws/output: init A1/A2 so that rgba64 ends with 0xffff in the absence of alpha input [02:07] <michaelni> Compn, ffplay code is different so i think its not affected [02:09] <Compn> ok , i wasnt sure how much code was shared between the tools [09:50] <nevcairiel> whats up with fate? it claims all machines havent sent data for 4 days [10:12] <cone-717> ffmpeg.git 03Michael Kostylev 07master:2768f7e4fef5: sparc: VIS mnemonics [10:12] <cone-717> ffmpeg.git 03Luca Barbato 07master:a5a0ef5e13a5: jpegls: return meaningful errors [10:12] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:b53d6a47c51a: Merge commit 'a5a0ef5e13a59ff53318a45d77c5624b23229c6f' [10:17] <michaelni> bcoudurier, the new server has different host keys, ssh will refuse to connect from clients unless people manually update their clients known_hosts [10:19] <michaelni> bcoudurier, if you are online & awake can you copy the old key over ? alternatively if you are afk, ill update my fate clients (and everyone else needs too) [10:37] <cone-717> ffmpeg.git 03Luca Barbato 07master:4a4107b48944: jpegls: factorize return paths [10:37] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:7d84bb2bcca8: Merge commit '4a4107b48944397c914aa39ee16a82fe44db8c4c' [11:12] <cone-717> ffmpeg.git 03Luca Barbato 07master:abad374909e6: jpegls: check the scan offset [11:12] <cone-717> ffmpeg.git 03Anton Khirnov 07master:bc8c1cdc7bfe: lavfi doxy: add a page for lavfi. [11:12] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:0718f0cc24bc: Merge commit 'bc8c1cdc7bfe7b1fe56df812aabdac3de2555cec' [11:15] <ubitux> so, should we update fate key? [11:20] <michaelni> ubitux, i updated my openbsd box and waiting to see if it works [11:24] <ubitux> just updated, seems to work [11:25] <cone-717> ffmpeg.git 03Anton Khirnov 07master:40020e171a35: doxy: add a section about versioning. [11:25] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:97200d89b9fc: Merge commit '40020e171a3549a2c0b65ce6f2649aec868872f2' [11:32] <nevcairiel> yeah seems fine, my fate also updated again [11:32] <nevcairiel> might want to send a mail to the ML to notify about the need to update the hostkey [11:35] <cone-717> ffmpeg.git 03Anton Khirnov 07master:f0cebf51ddfc: lavc doxy: add avcodec.h to the libavcodec group. [11:35] <cone-717> ffmpeg.git 03Anton Khirnov 07master:006ec64877fb: lavr doxy: add version.h to the lavr group [11:35] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:20f75a810c49: Merge commit '006ec64877fb638de4b15ae5fd969e22bd0f60ca' [11:36] <cone-717> ffmpeg.git 03Anton Khirnov 07master:e27f4b7cbf8a: doxy: don't sort struct members alphabetically [11:36] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:937af7f6d3e7: Merge commit 'e27f4b7cbf8a3fde313cfe4b3a1880bf5031e9ae' [11:57] <cone-717> ffmpeg.git 03Anton Khirnov 07master:b01f6041f426: lavfi: rename AVFilterFormats.format_count to nb_formats [11:57] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:74cb7ef85148: Merge commit 'b01f6041f4260fba053c2f96ce1611ea77e833a0' [12:28] <cone-717> ffmpeg.git 03Anton Khirnov 07master:c5ab6088ad7b: af_aformat: cosmetics, fix indentation [12:28] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:4742edd50e18: Merge commit 'c5ab6088ad7b0dac461fb6757b038f5260e638e5' [12:44] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:71b7427e9bdd: doc/fate: update server host key fingerprints [13:08] <cone-717> ffmpeg.git 03Anton Khirnov 07master:1dd35611433c: vf_format: do not use the AVFilterFormats internals. [13:08] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:dd2eaf57b760: Merge commit '1dd35611433c224e2231ed7ae26b5b560863a3cb' [13:15] <cone-717> ffmpeg.git 03Anton Khirnov 07master:d79bd6040599: lavfi: factorize freeing a link [13:15] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:fb8dde37e4b0: Merge commit 'd79bd6040599ef017dc45d5997fcc959ab1f71f6' [13:41] <cone-717> ffmpeg.git 03Anton Khirnov 07master:bf5b5d2b1561: avfilter.c: cosmetics, reformat [13:41] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:d56ced5fbf3c: Merge commit 'bf5b5d2b1561535cc013c12ab8033228bb0d0081' [13:55] <cone-717> ffmpeg.git 03Anton Khirnov 07master:7e2b15c094bb: avfilter: check a malloc in avfilter_link(). [13:55] <cone-717> ffmpeg.git 03Kostya Shishkov 07master:cebdedca57d9: prores: decode alpha plane when it's present [13:55] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:849bb62c2c9a: Merge commit 'cebdedca57d95834a8f8098c7b6a322a1163e26b' [14:38] <cone-717> ffmpeg.git 03Eli Friedman 07master:350914fc8024: Escape 130 (RPL) decoder [14:38] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:ed0aed09e8a1: Merge remote-tracking branch 'qatar/master' [14:55] <cone-717> ffmpeg.git 03Stefano Sabatini 07master:d91e44301c92: doc/filters: apply minor fixes to owdenoise documentation [15:15] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:e1efacf60bc8: sws/output: yuv2rgb_write() assert that alpha has the correct value in the tables [16:25] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:5e29e912327e: avcodec_align_dimensions2: add missing yuva formats [16:33] <cone-717> ffmpeg.git 03Piotr Bandurski 07master:7baef22fcad3: mov: set block_align for MACE 3:1 and MACE 6:1 [16:40] <burek> is there an opposite filter to channelsplit [16:44] <cehoyos> michaelni: Please merge proresdec2 from github, thank you! [16:48] <cone-717> ffmpeg.git 03Jean First 07master:7e570f027b9b: movenc: add video_track_timescale option [16:53] <Compn> are there still many differences between the decoders? [16:53] <Compn> just curious [16:53] <Compn> i guess since no one wants to work together, we are stuck with them :P [17:02] <cehoyos> Compn: I don't think you are right. [17:03] <cehoyos> See also https://ffmpeg.org/trac/ffmpeg/ticket/2575#comment:1 [17:27] <cone-717> ffmpeg.git 03Kostya Shishkov 07master:d13fa0e995ed: proresdec2: Support decoding transparency information. [18:08] <cone-717> ffmpeg.git 03Christophe Gisquet 07master:f49564c60759: fmtconvert: int32_t input to int32_to_float_fmul_scalar [18:15] <rexbron> cehoyos: I'm in if you want to work on the jpeg2000 issue. [18:15] <cehoyos> Which issue? [18:18] <Compn> people using ffmpeg to decode 4k stuff, nice :) [18:19] <rexbron> cehoyos: ticket 2586 [18:19] <cehoyos> For crashes, gdb output is unavoidable, sorry [18:20] <rexbron> cehoyos: are you able to decode the jpeg2000 frame on your end? [18:20] <cehoyos> Yes, but that is irrelevant: Crashes are always important, but without gdb output, we will not be able to fix it. [18:36] <cone-717> ffmpeg.git 03Michael Niedermayer 07master:9f88db5df141: avfilter/vf_deshake: check rx [18:52] <randomguy123> anybody can tell is there is a quick way to deinterleave NV12/NV21 into planar yuv? [18:52] <randomguy123> using some image_copy or similar without doing manual loop or using swscale? [19:07] <kierank> no [20:01] <randomguy123> kierank: thanks [22:33] <Daemon404> kierank, do people actually use 302m in real life? [22:33] <kierank> yes a lot [22:33] <Daemon404> ive only ever seen it in samples [22:33] <Daemon404> never real broadcast [22:33] <kierank> not to the home [22:33] <Daemon404> intermediate? [22:34] <kierank> yes between studio/control room etc [22:34] <Daemon404> right. [22:34] <kierank> i think mostly for dolby e [22:34] <Daemon404> as for dolby e, ive never even seen a sample ;) [22:34] <Daemon404> too pro for me [22:42] <nevcairiel> i think one sample has been flying around somewhere [22:44] <JEEB> there are plenty of samples it seems [22:44] <JEEB> I've seen such files go around P2P lately [22:44] <nevcairiel> considering that nobody can really play them :p [22:45] <JEEB> indeed [22:45] <nevcairiel> (unless they have the pro soft/hardware) [22:55] <kierank> people decode them with the pro software [22:55] <kierank> now that lavc has a demux [22:55] <kierank> but you have to decode to wav and re-encode [22:55] <kierank> you can't do realtime [23:02] <Daemon404> wow [23:07] <Compn> new fate box [23:07] <Compn> means everyone has to update fate settings probably [23:09] <nevcairiel> the ssh host key changed, so have to confirm the change once [23:10] <Daemon404> many ssh impls just will fail [23:10] <Daemon404> and you have to manually remove it [23:10] <Daemon404> or well [23:10] <Daemon404> i should say many openssh configs [23:10] <nevcairiel> putty asks nicely =p [23:10] <Daemon404> since everyone just uses openssh [23:10] <nevcairiel> with one big security warning [23:10] <Daemon404> yes [23:16] <cone-292> ffmpeg.git 03Lukasz Marek 07master:c86d3a54dedc: FTP protocol support [23:16] <llogan> what's the typeface in the current ffmpeg logo? [23:20] <Compn> herve should know [23:21] <Compn> if that logo is his... [23:22] <llogan> it is, AFAIK. i'll contact him, but i thought i should try being lazy first... [23:29] <llogan> we get some strange spam to the MLs: horse pellets and bedding, pigments, dog toys, etc. [23:31] <llogan> our resident trolls have been active lately. [23:44] <Compn> ffmpeg gains ftp support! better late than never ;) [23:54] <Compn> btw [23:54] <Compn> hmm [23:54] <cehoyos> Regarding Dolby E, I have heard rumours that there is a decoder: http://thread.gmane.org/gmane.comp.video.ffmpeg.issues/4526 [00:00] --- Sun May 19 2013

1 0

ffmpeg.log.20130518
by burek 19 May '13

19 May '13

[00:05] <Digit> anybody here ? [00:08] <JodaZ> Digit, do the mapping explicitly [00:09] <Digit> what does that mean... [00:09] <JodaZ> nvm [00:09] <Digit> :( [00:09] <Digit> how do i add the video file audio ? [00:13] <JodaZ> Digit, try -map "[out]" -map 0:1 [00:14] <Digit> before the -f thing ? [00:18] <Digit> wow it works :) i love you Jodaz :) [00:18] <Digit> i send you a lot of good energy :) thanks a lot [00:19] <Digit> i hope one day ... i'll understand those complex filters :P [00:21] <JodaZ> Digit, its not that hard, your input is numbered so you -map what you want in the output [00:21] <Digit> not that hard for you ;) [00:21] <JodaZ> Digit, also, see if -map 0:a works instead, because 0:1 is specific for that video [00:22] <Digit> well it works and i'm happy ^^ [00:22] <Digit> i am writing a gui for ffmpeg now it's really nice : ) [00:25] <Digit> lol now it seems it doesn't work with mpegts udp [00:25] <Digit> wwwwwwwow god lol [00:27] <Digit> Unable to find a suitable output format for 'filter_complex' gulp i start again ^^ [00:35] <Keshl> *God D: [00:40] <Digit> any idea why with udp as destination i get no image ? [00:58] <JodaZ> hmm, can i seek to a frame number ? like the 10th frame ? [01:14] <Digit> new problem :D i am sending my video with udp mpegts then i use another ffmpeg process to send it to internet with rtmp but sometimes i get only the sound and no image ! Any idea how i could fix it ? [01:14] <Digit> or what should i check [01:16] <Digit> oh i send the udp mpegts to local :) 127.0.0.1 [01:28] <Digit> and when i launch my 2 ffmpeg processes with the same command line i never get the image. ... [01:50] <Digit> any help ? [02:46] <Keshl> So, I have a video file that has a lot of motion. It ramps up from slow speed to high speed. I have taken minute-long samples from various parts of the footage and encoded them with the same parameters (ffmpeg -i foo.mp4 -vcodec libx264 -preset medium -crf 25 -movflags faststart -pix_fmt yuv420p OUT.mp4) and found that the beginning parts encode nearly twice as well (in terms of filesize reductions) as the later parts, and there comes a [02:46] <Keshl> steep point where it suddenly cuts off even as it gets faster. What I mean is, at the beginning of the video it compresses to about 27 megs. The middle goes to about 44 and the end is almost identical, only a few KB off. I have also noticed that, due to the high speed of the footage near the end, that individual frame quality makes much less of an impact. So, my question is, is there some way I can set -crf 18 at the beginning and ramp it [02:46] <Keshl> down to -crf 25 as the video goes on? Or something like that? [02:46] <Keshl> Or somehow tell it exactly at what second or frame to change the CRF? That'd work too. [02:47] <Keshl> Acrtually that'd work better. [02:48] <Kuukunen> Keshl: well, x264 command line has zones [02:48] <Keshl> Can'ya elaborate on that, oÉo? [02:48] <Kuukunen> but to me it sounds silly to do that just because the first part compresses better [02:50] <Kuukunen> I think the ratecontrol does a very good tradeoff for those [02:51] <Keshl> Not just the first part. [02:51] <Keshl> It makes more sense if I show part of the clips, one moment.. [02:52] <Keshl> Kuukunen: ftp://zone@connect.archgeekit.com:21918 -- Hug the "ffmpeg" folder. [02:54] <Keshl> Basically as the game progresses, the visuals constantly change between motion that's easily predicted to stuff that's much harder to predict and back again. At some point, because the game is moving so fast, CRF stays at a flat 33 according to ffmpeg's status output. The change in quality between visuals actually becomes noticable (in other words, CRF stops doing its job). I feel I really do need to help it. [02:54] <Digit> hi [02:54] <Keshl> This is especially noticable in the second video when it changes from yellowish to white. [02:54] <Digit> bcs you're using a too high value for crf [02:54] <Keshl> I'm using 25, actually. <É< [02:54] <Digit> too low [02:54] <Digit> i mean [02:55] <Keshl> oÉo Any lower and it all looks like crap. [02:55] <Digit> you need a lot of bandwidth to stream with that value [02:55] <Keshl> Not trying to stream, oÉo. [02:55] <Kuukunen> Keshl: you mean qp stays at 33? [02:55] <Keshl> Kuukunen: Yeah. [02:55] <Digit> why don't you try without crf [02:55] <Keshl> Digit: I'm trying to achieve what looks to be equal quality through the video, not a constant bitrate. [02:57] <t4nk765> Hi there, I'm trying to convert to mkv to mp4 and I get an error message of could not write header for output file #0. here's my pastebin: http://pastebin.com/cQxFvM0L [02:57] <Digit> i thought you were streaming it [02:57] <Keshl> Digit: Nope, oÉo. [02:58] <Keshl> t4nk765: Try this: ffmpeg -i (blah.mkv) -vcodec libx264 -preset ultrafast -crf 25 -movflags faststart -pix_fmt yuv420p -acodec copy OUT.mp4 [02:58] <Kuukunen> um [02:58] <Digit> are you sure the program can write to that directory [02:58] <Kuukunen> no you prolly don't want to reencode :P [02:58] <Keshl> OH. [02:58] <Keshl> My bad. [02:59] <Keshl> Kay. Ignore me on that. Do what Digit said. <.< [02:59] <Kuukunen> Keshl: did you use ultrafasat? [02:59] <Digit> ??? i said what ? [02:59] <Kuukunen> ultrafast* [02:59] <Keshl> Kuukunen: No, medium. [02:59] <Digit> ultrafast is shitty [02:59] <Keshl> I just saud ultrfast to him, oÉo. [02:59] <llogan> t4nk765: flac can't go into mp4 container [02:59] <t4nk765> Keshl:what does crf and movflags and pix_fmt do? [02:59] <Keshl> I can't think of a good way to explain them. [03:00] <Digit> it's just about your codec parameters [03:00] <Keshl> Thankye, oÉo. [03:00] <llogan> oops...i meant https://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [03:00] <Kuukunen> Keshl: oh yea, just google: x264 zone [03:00] <Kuukunen> to get info [03:00] <Keshl> Danks, oÉo. [03:01] <Keshl> Though since the transfers are "almost" done I'll wait for your advice first in case there's something better I can try. [03:01] <Digit> - options: is not correct [03:01] <Digit> Keshl [03:01] <Keshl> oÉo? Where? [03:01] <llogan> t4nk765: see link i just provided for crf. movflags allows your video to begin playback before it is completely downloaded by client (as in viewing your video in web browser). [03:02] <llogan> "-movflags faststart" specifically [03:02] <Digit> [mp4 @ 0xa01f920] track 1: could not find tag, codec not currently supported in container [03:03] <Keshl> Digit: Are you the one downloading my videos? You know none are done yet, right? [03:03] <Keshl> Well, now the first one is. [03:03] <Digit> i'm downloading nothing [03:03] <Keshl> oÉo. Nevermind then. <É<' [03:03] <t4nk765> I I see. thanks for that. although the command fails as someone has pointed out I can't put flac into mp4. what is a good alternative audio you can suggest? [03:03] <Keshl> t4nk765: Don't supply -acodec, it'll handle it its self. [03:03] <Digit> mp3 [03:04] <llogan> t4nk765: most common is AAC [03:04] <llogan> https://ffmpeg.org/trac/ffmpeg/wiki/AACEncodingGuide [03:04] <llogan> libfdk_aac is usually recommended now [03:04] <Digit> i got a question too :) [03:04] <t4nk765> should I specify AAC or just let the program take its course by omitting acodec like keshl has said? [03:05] <Digit> one ffmpeg process sends a video to udp 127.0.0.1 and another ffmpeg process use it as a source and sends it to a server using Rtmp [03:06] <llogan> t4nk765: selection of encoder can depend on your ffmpeg configuration, and not all encoders are equal, so it is best not to omit [03:06] <Digit> the problem is that i need to wait for some seconds before i launch the second process or i get only audio and no image.... [03:07] <Digit> is there a way to fix that ? [03:07] <Kuukunen> Keshl: I was downloading them [03:07] <Keshl> Kuukunen: oÉo. I figured it was you or Digit. Why'd you cancel the other two? They was so close and I only has 200kb upload bandwidth .É. [03:08] <Kuukunen> Keshl: personally I think the quality is pretty similar among the three... [03:08] <Kuukunen> have you been looking at still pictures? [03:08] <llogan> t4nk765: yuv420p http://en.wikipedia.org/wiki/Chroma_subsampling [03:08] <Keshl> Watch the ship. >w> And nu, oÉo. [03:08] <t4nk765> alright. thanks for your help :D [03:09] <Kuukunen> Keshl: also, try veryslow [03:09] <Kuukunen> usually it's fast enough :P [03:09] <llogan> most players only handle YUV 4:2:0 planar when you encode H.264 video [03:09] <Keshl> Kuukunen: Why not placebo then? XD [03:09] <Digit> placebo sux [03:09] <Keshl> The problem is, normally, I actually use placebo. [03:09] <Keshl> But then playback only works if you're running an nVidia Titan or something. [03:09] <llogan> placebo is a joke and a waste of time [03:09] <Keshl> <.< [03:09] <Kuukunen> Keshl: placebo isn't worth it [03:10] <Kuukunen> that's why it's called placebo [03:10] <Keshl> oÉo.. Why speciffically? .É. [03:10] <Kuukunen> Keshl: uses lots more time for no real gain [03:10] <Keshl> ...Holly, veryslow's epicier. o_o [03:11] <Kuukunen> Keshl: also, which direction did you even want to tweak the faster half? [03:11] <llogan> less than 1% over veryslow [03:12] <Digit> any idea for my problem ? [03:12] <llogan> sleep 3; ffmpeg ... [03:12] <Digit> and another question is it possible to use normal filter with a complex filter ? [03:12] <Keshl> Kuukunen: It's not the faster part sof the game that get me, but see how in the beginning the course just flashes? That part I'm fine with being low quality (Except I want the first few seconds to be really high quality, just cuz.), but I want the stuff with more detailed visualizers (like the yellow part in the second video) to be higher quality. [03:13] <llogan> Digit: what do you mean by "normal" filter? [03:13] <Digit> a not complex one :) [03:13] <Keshl> And then at the very end (zone 75 and past) there's no visuals but it's so fast I'd like to turn the quality up again, so. [03:13] <Digit> -vf [03:13] <Digit> use the tune grain [03:13] <Kuukunen> Keshl: why do you want to force more bitrate in places where you can't see anything because it's moving so fast? >_> [03:14] <Keshl> Kuukunen: Because OCD> [03:14] <Keshl> And because almost nobody gets that far. [03:14] <Keshl> <.< [03:14] <Keshl> It's very rare to see people pass zone 75. So. [03:14] <llogan> -vf is used for a simle filtergraph: one input and one output. -filter_complex is for complex filtergraphs, "which cannot be described as simply a linear processing chain applied to one stream" such as more than one input and or output [03:15] <Kuukunen> Keshl: use qp then :P [03:15] <llogan> http://ffmpeg.org/ffmpeg.html#Complex-filtergraphs [03:15] <Digit> so i cannot use both ? [03:15] <Keshl> Kuukunen: ....Oh duh. Okay, thanks. -É- [03:15] <Digit> at the same time i mean [03:15] Action: Keshl purrpaws at. [03:15] <llogan> Digit: no. there is no reason to [03:16] <Kuukunen> Keshl: but what you end up with in the end, imho, is bigger file with worse quality :P [03:16] <Digit> well the reason is taht i don't know how to add my "normal" filters to the complex one :D [03:16] <llogan> what are you trying to do? [03:16] <Keshl> Kuukunen: Okay, then what do you recommend I try, oÉo? Just stick to CRF? [03:17] <Kuukunen> yea [03:17] <Keshl> Even though qp stays at 33 throughout? [03:17] <Keshl> oÉo [03:17] <Kuukunen> if it's not supposed to be looked at still frame by still frame, just use xrf [03:17] <Kuukunen> crf* [03:17] <Digit> well i use now a complex filter to overlay an animated image it was not working with the normal filter [03:17] <Digit> but i had a normal filter for cropping and scaling [03:17] <llogan> Keshl: http://slhck.info/crf.html [03:17] <Kuukunen> don't look at the qp, look at the quality :P [03:17] <Keshl> oÉo... [03:18] <sacarasc> Owo? [03:18] <Kuukunen> Keshl: if it's not enough, use higher quality crf [03:18] <Keshl> It's :3 rotated sideways, oÉo. [03:18] <llogan> i think that's a ball-sack in the middle [03:18] <Keshl> See http://4.bp.blogspot.com/-bMlNRnlkteU/UGU1qZzuZzI/AAAAAAAAAek/Vmn_3mor1oo/s… [03:18] <Kuukunen> Keshl: 25 is pretty low quality [03:19] <Keshl> Kuukunen: 18 has no noticable improvements and more than doubles the filesize. <É< [03:19] <Keshl> Doesn't iomprove till I'm near 8 or lower. D: [03:19] <Keshl> *improve [03:19] <Keshl> llogan: No. These are balls: ~ [03:19] <Keshl> (Actual unicode character!) [03:19] <llogan> heh [03:19] <Keshl> So there. [03:19] <Kuukunen> Keshl: I don't believe you about not getting improvement :P [03:20] <Keshl> Kuukunen: Okay, I'll toss the raw footage at you. It's about 200 megs. Which one you want, oÉo? [03:20] <Kuukunen> whichever :P [03:21] <llogan> Keshl: did you read the link I provided? [03:22] <Digit> http://pastebin.com/sRDP5ZJ0 [03:23] <Digit> and any idea for my first question ? [03:23] <Keshl> llogan: Multitasking, getting around to it, oÉo. [03:23] <Digit> my program is already multistaking the 2 processes [03:24] <Digit> but if the second ffmpeg is not launched at least 6 sec after the first one i get no image [03:24] <Digit> only the sound is streaming [03:26] <Keshl> Kuukunen: Kay maybe I am obsessing over numbers now that I look at it closely.. Annnyway, same link, ftp://zone@connect.archgeekit.com:21918 , just in the "RAW" folder. oÉo. [03:26] <Digit> so i wonder how to make my program wait for something from the first ffmpeg to know it's ready to launch the second ffmpeg [03:27] <Digit> and what does it need to wait for ... [03:27] <Keshl> Just tell it to wait 10 seconds, oÉo? [03:28] <Keshl> Kuukunen: Also, -veryslow compared to -medium. The results are.. Really, weird. [03:28] <Keshl> But I think they're important to mention. [03:28] <Keshl> On the first video, it goes from 27,282 KB to 23,515 KB. [03:28] <Keshl> On the second.. Absoutely, no change. [03:28] <Kuukunen> sounds reasonable [03:28] <Digit> slower is tha best [03:28] <Keshl> Still doing the third. [03:29] <Keshl> And I'm suire I did -veryslow, not accidently re-ran -medium. [03:30] <Keshl> And I wanna point out, the second one has the most "intense" visualizer of the three videos. [03:30] <Kuukunen> Keshl: veryslow uses different ways of doing things... all of those methods are not as effective on fast paced video [03:30] <Keshl> (As in, the footage in general is the least predictable) [03:30] <llogan> Digit: you want to scale and crop the output of the overlay, or the input(s) of the overlay? [03:30] <Keshl> Third is smaller, but I accidently stopped it a few seconds before finishing. <.< [03:31] <Keshl> Okay, trying -slow... [03:31] <Digit> i want to scale the input video [03:31] <Kuukunen> Keshl: veryslow is better [03:31] <Kuukunen> it's just slower than slower [03:31] Action: Keshl tries placebo! [03:31] Action: Keshl waves a flag! [03:32] <llogan> Digit: which input? [03:32] <Digit> before it gets compressed with x264 [03:32] <sacarasc> Prepare for it to take 10 times as long for almost no improvement, Keshl. [03:32] <Digit> test.avai [03:32] <Digit> test.avi [03:32] <Keshl> sacarasc: Well, veryslow took about four times as long as medium and the filesize ws identical. <.< [03:32] <Keshl> *was [03:32] <Kuukunen> Keshl: what about quality? [03:32] <Keshl> Worse. [03:32] <Keshl> Honestly. [03:33] <Kuukunen> ok, I can try that one too :P [03:33] <Keshl> I've noticed that before. Medium always looks better than the slows to me, then placebo suddenly looks awesome. [03:33] <Keshl> (And is still smaller than medium) [03:33] <Digit> lol slower is a lot better [03:33] <Digit> or use tune film [03:33] <Keshl> Okay, okay, how about I torrent these raw videos instead and you can all play with them? [03:33] <Keshl> XD [03:35] <Keshl> -slow came out with about 1.5 megs smaller than -veryslow, and looks better. [03:36] <Keshl> Trying veryslow next, then placebo, cuz reasons.. [03:36] <llogan> Digit: -filter_complex "[0:v]scale=iw/2:-1,crop=iw:ih-20[foo];[foo][1:v]overlay[out]" [03:36] <llogan> the [out] is superfluous in this case, probably [03:37] <Digit> but my image is not there any more ?! [03:37] <Digit> i mean my animated gif [03:38] <llogan> console output! [03:38] <Digit> you know the one in my pastebin [03:38] <llogan> my crystal ball is broken from overuse [03:38] <Digit> http://pastebin.com/sRDP5ZJ0 [03:38] <llogan> ok. i'm done trying to help you. [03:39] <Digit> ?? [03:39] <llogan> you never include the ffmpeg output [03:39] <llogan> even after being asked repeatedly over the day [03:39] <Digit> yes but it's not for a problem, it's bcs i don't know how modify the command [03:39] <llogan> i gave you the filter_complex. what else do you need? [03:40] <Digit> well i leave you alone you seem really upset thanks anyway... [03:41] <Keshl> slower, about 1 meg smaller than slow, worse quality. [04:10] <Keshl> Placebo, biggest benefit so far at 3 megs. Quality is best. [04:10] <Keshl> So there. [04:11] <Keshl> (Too bad you need a $1,000 GPU to play it..) [04:13] <Digit> i want to thanks again, now my animated gif is displayed correctly on my stream, i learned a lot here, see you next time : ) [04:14] <klaxa> <Keshl> (Too bad you need a $1,000 GPU to play it..) [04:14] <klaxa> haha what [04:23] <Keshl> klaxa: Playing 720p at 60 FPS is painful. [04:23] <Keshl> <.< [04:23] <Keshl> Gets choppy on Placebo unless you've got seriously powerful hardware. [04:24] <Keshl> klaxa: Hug ftp://zone@connect.archgeekit.com:21918 and poke ffmpeg/OUTx264PLACEBO_2.mp4 [04:46] <Kuukunen> Keshl: uploaded some screenshots: http://kuukunen.net/misc/zon/ [04:47] <Kuukunen> every two are a pair [04:47] <Kuukunen> like, a<->b, c<->d, etc [04:47] <Kuukunen> which ones do you think are best looking? :V [05:19] <Keshl> Kuukunen: A, D, F, G, J , oÉo. [05:34] <Keshl> .... So I finally read that link llogan provided. [05:35] <Keshl> Changed crf to 23. Everything's happy-squee now. [05:35] <Keshl> ... ..... So yeah. [05:35] <Keshl> ^É^' [05:35] <llogan> 23 is also the default. [05:37] <Keshl> Yes, I read that bit.. XD [07:03] <Keshl> .É. What'd I break? D: [07:03] <Keshl> Oh right. [07:03] <Keshl> What'd I break? D: http://pastebin.com/Kd6vxUxS [07:07] <Keshl> Wow, /already/, Google indexed that. [07:13] <angelsl> hi, how do i get ffmpeg to encode a PNG + a .WAV to a x264 and ending the video when the .wav ends? [07:13] <angelsl> right now i have ffmpeg -loop 1 -i f.png -i TITLE.wav -c:v -tune:stillimage -r 1 test.mp4 [07:13] <angelsl> but that never ends, probably because of -loop 1 [07:17] <Keshl> Gota, was just outdated, oÉo. [07:17] <Keshl> angelsl: Supply -t (Total length of video in seconds) [07:18] <angelsl> there's no way to get ffmpeg to stop when one input ends? [07:20] <Keshl> angelsl: Dunno, oÉo. [07:20] <angelsl> found it, -shortest [07:20] <angelsl> thanks :) [07:20] <Keshl> Shiny -É- -Paws at.- [07:20] <Keshl> Welcomes opwo [07:20] <Keshl> *oÉo [07:21] <Keshl> What's the latest stable, official release for ffmpeg, oÉo? Like, the actual version number. Is it still 0.9.2? [07:24] Action: Mavrik points at topic. [07:24] <Keshl> Oh. [07:24] Action: Keshl expands. [07:24] <Keshl> .... Ohhhhh [07:24] <Keshl> Danks ^É^ [07:26] <Keshl> .É.! [07:26] <Keshl> http://pastebin.com/Lcw0rB1K Whus I do? D: [07:26] <Keshl> (It does keep going, but what's it mean and am I gunna die now? D: ) [07:27] <Mavrik> the decoder is having problems with input stream [07:27] <Mavrik> either its broken, you're running out of bandwidth, etc. [07:28] <Keshl> Must be broken, then. .É. [07:29] <Keshl> Most that'll happen is probably some "green frames", right? Or something like that? oÉo? [07:40] <angelsl> ffmpeg always overshoots -shortest when i'm encoding to h264 [07:40] <angelsl> but everything else works fine [07:54] <lnostdal> hi guys, you probably get this question all the time, but the official(?) ubuntu ppa seems out of date .. and i prefer the "real" ffmpeg to the fork included in ubuntu .. is there another "semi-official" ppa out there people generally use? [08:08] <retard> lnostdal: people use a static binary build [08:10] <retard> burek gracefully provides them at http://ffmpeg.gusari.org/static/ [08:36] <lnostdal> doesn't seem to work with pulse / also .. hm [08:36] <lnostdal> alsa* [08:38] <lnostdal> i'll just build from source i guess [09:31] <lnostdal> "opus not found" .. not sure how to build (configure) this against a local static opus.. [09:52] <burek> lnostdal --with-opus=path perhaps? [09:55] <burek> --enable-libopus [11:10] <doub> is it possible to do basic video editing with ffmpeg? basically all i need is only convert some specified segments of the input file [11:25] <retard> doub: yes [11:32] <doub> i've found the -t and -ss option, is that the right method to do that? can i specify several segments? [12:31] <Digit> hi :o) [13:06] <retard> doub: no, but you can do it in stages [13:06] <retard> extract the relevant clips, then combine them again [13:07] <Digit> do what ? [13:09] <retard> he wants to [13:10] <retard> or she, though doub sounds male to me for some reason [13:23] <Digit> when ffmpeg read some video files is it possible to choose deblocking or some postprocessing ? [13:29] <retard> doub: you might have missed my reply due to timing out, but split the task into steps, extract each clip then combine [13:29] <retard> or use a more high level tool [14:05] <Digit> how do i become a moderator ? :) [14:06] <esing> Hi, would you recommend to capture a 15-20 years old vhs cassette (PAL, interlaced, composite, encoding with mpeg-2@ a bitrate of 9800kbps) with uyvy or yv12(=yuv420) or even another color space. I have no clue which one to pick [14:12] <Digit> it depends more where you will use it , i think [14:14] <Digit> esing read this http://www.spearsandmunsil.com/portfolio/choosing-a-color-space-2/ [14:15] <esing> Thanks Digit [14:15] <Digit> you're welcome [14:34] <luc4_linux> Hi! I have an application which muxes h264 into mp4. Seemed to work. Now I ported to another system but I'm getting: [NULL @ 0x13e67a0] Unable to find a suitable output format for 'out.mp4'. I suppose something is not compiled in this version of ffmpeg. Anybody who can give me a hint on what maybe? [14:35] <Digit> PASTE TO http://ffmpeg.pastebin.com/ ! If you are asking about CLI usage, paste the complete command and output to the paste bin. [14:37] <luc4_linux> pastebin one single line? [14:37] <Digit> what do you mean by "an application" and "ported to another system" , that's not sharp enough [14:37] <Digit> not one single line all the ouput too [14:38] <luc4_linux> Digit: I built my code under linux. Seemed to work. Now I want to compile for android, linking to ffmpeg compiled for android. I compiled the libs for android but I'm getting that. And by that, I mean only that. Nothing more. [14:39] <Digit> well if you dont pastebin all the output you'll get probably no help here [14:39] <luc4_linux> Digit: ...? That is all the output. [14:40] <Digit> the complete command and output [14:40] <luc4_linux> Digit: command? not using the ffmpeg command... [14:40] <luc4_linux> Digit: I'm using libavformat and libavcodec. [14:40] <JEEB> :) [14:40] <Digit> :) [14:40] <luc4_linux> Digit: and again, that is all I get from the libs. The output is a single line. [14:41] <JEEB> luc4_linux, well it's saying that it can't guess what you want from 'out.mp4' [14:41] <JEEB> do you have the mov/mp4 muxer in there? [14:41] <JEEB> in the ffmpeg you're linking against [14:41] <JEEB> tried manually specifying the format? [14:42] <luc4_linux> JEEB: I suppose yes, I enabled that in cofigure after --disable-everything. Do I need to enable something else? [14:43] <JEEB> are you sure you have file output as such enabled, too? [14:43] <JEEB> --disable-everything can be rather limited :P [14:44] <Digit> :D [14:44] <JEEB> anyways, try with a standard ffmpeg first, and then move to a more limited one after making sure it works [14:44] <luc4_linux> JEEB: ah ok right, I'll try that, thanks! [14:46] <luc4_linux> JEEB: Before I compiled a slightly different version, and got "Protocol not found." when trying to open the output file. Do you know what this means? [14:48] <JEEB> not completely sure because I have no idea what exactly you did nor do I remember the internals of libavformat, but I have a hunch that that was you missing the whole file writing thing :P [14:48] <JEEB> but... go try out that same thing without everything disabled first [15:12] <esing> Digit, Hmm.. to me the safest way seems to pick 4:2:2. From what I read from your link, it looks like most players do good conversions from 4:2:2. Did you read the article too? [15:16] <Kuukunen> Keshl: lol, you might want to look at a and b again, because I think that's completely clear :P [15:17] <Kuukunen> Keshl: the last four you picked, in order, veryslow over slow, veryslow over medium, veryslow over slow, veryslow over medium [15:17] <Kuukunen> which makes sense >_> [15:19] <Kuukunen> (A is veryslow crf 25 and B is veryslow crf 18) [15:21] <esing> Um.. this confuses .... you see in following link that mpeg2 DVD is restricted to 4:2:0 http://en.wikipedia.org/wiki/MPEG-2#DVD-Video My question: Does that mean that if I encode something to mpeg-2 that I am not restricted to 4:2:0, but can also use 4:2:2 and play the encoded mpeg2 video from the computer or usb stick on a display in 4:2:2 and only if I create a DVD iso from the mpeg2 encode that only _then_ Iam restricted to the colorspace 4:2:0 ? [15:24] <Kuukunen> esing: mpeg2 itself allows 4:2:2 and 4:4:4 [15:24] <esing> Good to know thanks [15:25] <Digit> anyway why would you use something else than 4:2:2 with a poor quality vhs [15:25] <Digit> you will not even see the difference [15:25] <Kuukunen> esing: mpeg2 also allows resolulutions like 1920x1152, but it doesn't make much sense for DVDs to allow that :P [15:26] <Digit> for vhs source i would not even use more than 640x [15:26] <esing> Kuukunen, If I have a 4:2:2 mpeg encode and I want to create a DVD from it (with dvdauthor) ... do I then have to care about the conversion from 4:2:2 to 4:2:0 , or does dvdauthor does that job [15:26] <Kuukunen> umh :P [15:26] <Digit> you have to care about it if you want to use it somewhere like on justin tv [15:27] <Kuukunen> esing: you're making it a DVD? [15:27] <esing> Kuukunen, I make it primarily an mpeg2 encode storing it on my hdd, but later I will make some dvd's out of it [15:27] <Kuukunen> if you want to store it on your hdd, why on earth would you use mpeg2? :P [15:28] <Digit> to move it to dvd after that [15:28] <esing> Yes, because I think it is easier to put mpeg2 to dvd [15:28] <Kuukunen> Digit: yea, but if he's planning on making it a 4:2:2 mpeg2, he has to reencode anyway :P [15:28] <Digit> life is life :D [15:28] <esing> So the wikipedia information is right, I can't put that 4:2:2 on a dvd hmm [15:29] <Digit> wikipedia is always right :D [15:29] <esing> Heh :) [15:30] <Digit> anyway i use files on my dvd so i can use h264 compression [15:30] <Kuukunen> esing: I'd just capture it losslessly (or near losslessly) then filter it and encode it to H.264 for storage... [15:30] <Kuukunen> esing: if you know you want it on DVD, encode it from the filtered lossless to the DVD format [15:31] <Kuukunen> and yea, of course you can store files on DVDs [15:31] <Kuukunen> but I assumed that you want an actual playable DVD [15:31] <Digit> well now every players can read files ... [15:31] <Digit> no ? [15:33] <esing> Iam not sure if it is worth to encode it to H.264 for storage in comparision to mpeg- [15:33] <esing> mpeg2* [15:33] <Kuukunen> esing: I thought you wanted higher quality :P [15:34] <esing> Kuukunen, I think that mpeg2 with the right options should make nearly no difference for a vhs tape compared to h264 [15:34] <Digit> are you kidding me ? [15:34] <Kuukunen> esing: well, of course you can make a high quality mpeg-2 [15:34] <Digit> with a vhs ??? never [15:34] <Kuukunen> esing: it just means it's four times bigger than the h.264 [15:34] <Digit> that's right [15:35] <Kuukunen> (not sure if it's four, it might be something between two and ten) [15:35] <esing> Ye, that's fine.. so I would neither have to store the lossless capturing [15:36] <esing> Thus I think the sane way (for the mpeg2 to encode) is to choose 4:2:0, so I can make a dvd out of the stored videos in a minute [15:36] <esing> That is kind of compromiss because if the vhs would be in 4:1:1 color space I'd loose a little bit quality I guess [15:36] <Kuukunen> using mpeg2 files for video storage in 2013 sounds plenty non-sane to me :P [15:37] <esing> heh :D [15:37] <esing> Well we talk about 15-20 years old vhs cassettes :) [15:38] <Kuukunen> yea [15:38] <Kuukunen> but even still [15:38] <Digit> well compress it with h264 if you don't want to move it now to dvd and then you can still transcode it later anyway [15:38] <klaxa> Kuukunen: ever heard of blurays [15:38] <Digit> that's it [15:38] <klaxa> also mpeg2 container or video format? [15:39] <klaxa> well if we're talking DVD probably codec [15:39] <Kuukunen> klaxa: why would you use mpeg2 on a blu-ray :P [15:39] <esing> klaxa, video format [15:39] <klaxa> TV is still widely broadcasted in mpeg2 [15:39] <Kuukunen> it is [15:39] <klaxa> but yeah for storage it's barely used [15:40] <Kuukunen> but that's because of legacy [15:40] <klaxa> or rather it should be barely used [15:41] <Kuukunen> yea, otoh, the difference between the size becomes smaller when you get to very high quality [15:42] <Kuukunen> but still... no mpeg2 when you have a choice ... [15:42] <Digit> next question pls :) [15:43] <Digit> is it possible to modifiy in real time some options, without launching ffmpeg again with the new options [15:43] <Digit> i mean when you're streaming [15:48] <luc4_linux> Hi! When using libavcodec I'm getting: Failed to find codec h264. I added --enable-libx264 in configure. Is there anything else I should add to get rid of this? [15:53] <Digit> for debian ? [15:54] <luc4_linux> Digit: on ubuntu the version I built is fine, the version I'm building for android seems to give that. [15:55] <Digit> oh yes for android ... [15:58] <luc4_linux> I see libavcodec/libx264.o is built... so weird it is not found... [15:58] <Kuukunen> luc4_linux: you mean it compiles and you get that when you try to run it? [15:59] <luc4_linux> Kuukunen: you mean libavcodec/libx264.o? Yes, I see it in the output of the build of ffmpeg. [15:59] <Digit> you want to build a ffmpeg for android that's right ? [16:00] <Kuukunen> luc4_linux: no I mean, when do you get the error? [16:00] <Kuukunen> are you trying to run it or build it with codec "h264" [16:00] <Digit> i got a nice tuorial to build one [16:00] <Digit> http://www.roman10.net/how-to-build-ffmpeg-for-android/ [16:02] <luc4_linux> Kuukunen: I'm trying to run my application using libavformat and libavcodec and I get that. I suppose I'm missing something in the configure... [16:02] <luc4_linux> Digit: ffmpeg builds fine for android... the problem seems to be the configure options... [16:03] <Kuukunen> luc4_linux: it might be because you should use "libx264" instead of "h264" [16:03] <luc4_linux> unless something is wrong with libx264... [16:03] <luc4_linux> but I added --enable-libx264 [16:03] <Kuukunen> luc4_linux: are you decoding or encoding? :P [16:04] <luc4_linux> Kuukunen: I know it sounds weird... but none :-) I'm only trying to mux. But that seems to be necessary anyway... [16:04] <Kuukunen> umh :S [16:04] <luc4_linux> Kuukunen: anyway, maybe something is not entirely ok with my code, but works on ubuntu [16:04] <Paranoialmaniac> luc4_linux: libx264 is not needed for muxing [16:06] <luc4_linux> Paranoialmaniac: I guessed that :-) maybe I've done something not required in my code which causes it to be needed. Anyway, for the moment it is ok anyway... just don't know why it is not found... [16:06] <spumco> o/ [16:07] <spumco> is this the right place to ask about ffmpeg audio channel mapping? [16:09] <Digit> ask and pastebin ouput [16:10] <Kuukunen> luc4_linux: maybe it's because you have libx264 but not h264? [16:11] <spumco> basically i have an uncompressed quicktime file that contains video and 16 channels of audio. i want to take the first 2 channels of audio and map them to left/right. the video converts just fine into h264 every time, but i always get blank audio. i've been trying to work out what the map_channel options should be. [16:11] <elkng> I have content of entire DVD copied, in directory VIDEO_TS, and it contains few files like: VIDEO_TS.BUP, VIDEO_TS.IFO, VTS_01_1.VOB and others, do I need to use some special command to convert that DVD ? or I just need to use regular encoding command on *.VOB files and delete all others ? [16:11] <luc4_linux> Kuukunen: mmh... I'll try to add --enable-decoder=h264 then. thanks [16:12] <JEEB> also I think the annex b parser was in libavformat :P [16:12] <JEEB> as a format [16:14] <Digit> spumco i think you just have to use a complex filter [16:16] <spumco> do you mean -filter_complex? i'd seen that but i thought it was possible to use map_channel as a selector? i've been using "-map_channel 0.1.0 -map_channel 0.1.1" to select them. the output is listed as stereo but the audio is always blank. i know channels 0 and 1 in stream 1 contain my audio for sure [16:17] <Digit> pastebin your command and output [16:17] <luc4_linux> I don't know what made it work, but now it works. Thanks guys! [16:17] <luc4_linux> Clearly something was wrong in my configure command. [16:18] <Digit> share it so we can say the solution to the next one [16:19] <JEEB> "don't try to disable too many things in your configuration right off the bat" [16:19] <JEEB> you always forget something :P [16:19] <spumco> is this correct? not used pastebin before - http://pastebin.com/PHqdfRcd [16:20] <Digit> spumco see if some of those examples could help you : http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20use%20-map%20option [16:21] <spumco> none of those options show channel selection though, they're all about stream level selection [16:23] <spumco> i notice there's a complaint part way down from something called pan about deprecated syntax [16:23] <spumco> and then it goes on to say "Pure channel mapping detected" [16:25] <Digit> yes i was reading that too ... [16:26] <Digit> pure means it uses losless [16:27] <spumco> which would tie up with what vlc reckons the output mp4 contains - mpeg aac mp4a stereo 48KHz [16:27] <spumco> input is 16 channel 48KHz [16:27] <Digit> well i can't find an answer to help you [16:28] <Digit> may be another guy here is smarter [16:28] <spumco> hehe, thank you for trying [16:28] <spumco> i've been fighting it all morning [16:29] <spumco> i used mplayer to dump the raw audio then audacity to work out which channel was which [16:29] <Digit> i fighted 2 weeks to find a solution to my "problem" :P [16:29] <spumco> that's how i found channels 0 and 1 had my audio [16:30] <Digit> so you just get no audio at all ? [16:30] <Digit> did you try to play it with vlc ? [16:30] <spumco> no no, that was just me trying to work out what format the audio was [16:30] <spumco> there's no obvious documentation on the format from the product [16:31] <spumco> but i know it's 48 khz, 16 channels, pcm, little endian [16:31] <spumco> audacity will read the raw file as that and it works perfectly [16:31] <Digit> gulp [16:31] <spumco> i just can't get ffmpeg to take the 2 channels from stream 1 and shove them into aac [16:32] <Digit> the output audio seems to be fine... [16:32] <spumco> i was trying to avoid seperating the 2 to avoid sync problems [16:33] <spumco> i guess i can use ffmpeg as a multiplexer if i can split the audio and junk the extra streams [16:33] <spumco> but that's a guess [16:33] <spumco> i'm sure i'm just screwing up the command line options somehow though [16:35] <Digit> can't you really try with a filter complex ? [16:36] <spumco> i'll go look at that now :) [16:38] <burek> spumco, did you see https://ffmpeg.org/ffmpeg-filters.html#channelmap [16:38] <spumco> interestingly quicktime doesn't know how to map the audio - it just mixes all the channels to mono [16:39] <spumco> yes, i've been looking at that [16:40] <burek> btw, you can extract all 16 channels into separate audio streams, to examine them even further with https://ffmpeg.org/ffmpeg-filters.html#channelsplit [16:40] <Digit> the syntax your used is not the one i know [16:41] <Kuukunen> JEEB: ./configure --help|grep -o -- '--enable-[a-z0-9-]*\s'|xargs ./configure [16:41] <Digit> i can't see something like -c copy in your command [16:41] <spumco> burek, i'll try that to see if ffmpeg can dump the audio. i did it with mplayer originally to get a single large file. [16:42] <JEEB> Kuukunen, the problem generally is the other way :P [16:42] <Kuukunen> JEEB: I mean, that's the solution! [16:42] <JEEB> --disable-everything and then they think they have enabled enough while they haven't [16:42] <JEEB> -_- [16:42] <Kuukunen> enable averything \o/ [16:43] <Kuukunen> I'd say that's often a problem too... when people try to enable everything :P [16:43] <Kuukunen> and then complain about dependencies, or something [16:44] <burek> spumco, also, you kinda didn't use the correct syntax [16:44] <burek> -map_channel [input_file_id.stream_specifier.channel_id|-1][:output_file_id.stream_specifier] [16:45] <spumco> yer, but i thought the 2nd part was optional. i tried with "-map_channel 0.1.0:0.1 -map_channel 0.1.1:0.1" but the output audio was still blank :( [16:45] <burek> hmh, i see that -map_channel is probably a shortcut for pan filter [16:45] <burek> let me check something [16:45] <spumco> i think you are correct [16:46] <spumco> i did try going straight to pan but dug myself a different kind of hole [16:46] <Digit> use something like this : ffmpeg -i input.mkv -map 0:1 -map 0:2 -c copy output.mkv [16:46] <burek> https://ffmpeg.org/ffmpeg-all.html#Remapping-examples [16:49] <burek> spumco, so try pan filter from the upper link, to get something like: ffmpeg -probesize 500000000 -i sample-uncompressed-lr-audio.mov -vcodec libx264 -vf "pan=..." output.mp4 [16:50] <burek> you might also try to specify "-c:a copy" since you are using a lossless remapping [16:50] <burek> ffmpeg -probesize 500000000 -i sample-uncompressed-lr-audio.mov -vcodec libx264 -vf "pan=..." -acodec copy output.mp4 [16:51] <spumco> -map 0:0 -map 0:1 gives an mkv file with working video but 16 channels of audio [16:51] <spumco> i don't know how to tell vlc to play channels 0 and 1 so it's silent [16:51] <spumco> can you put pcm into an mp4 container? i thought it had to be aac? [16:58] <Digit> i was sure it would leave only 2 channels ... [16:58] <spumco> i think at that level, they're streams [16:58] <spumco> so 0:1 is file 0, stream 1 [16:58] <spumco> to get to channels you have to go another level down [16:59] <spumco> so 0.1.0 is channel 0 in stream 1 of file 0 [16:59] <spumco> if i use "-map_channel 0.1.0 -map_channel 0.1.1" it works out i want stero [16:59] <spumco> stereo [16:59] <spumco> i'm just not sure why it's blank, heh [17:00] <spumco> i don't know how to specify 16 channels in channelsplit either [17:03] <spumco> i'm not sure it maps to any standard channel layout [17:03] <spumco> it's just 16 channels :( [17:11] <Digit> spumco i don't understand what you need bcs it works for me with -map ... [17:12] <Digit> did you use the info give by ffmpeg to use the correct channels numbers... [17:12] <Digit> may be i understand nothing ... [17:12] <spumco> what have you tried? [17:12] <spumco> and what command line did you use? [17:13] <spumco> <- reading about pan options [17:13] <Digit> i tried my command line with -map to remove one channel in a mpeg and it seems to work [17:14] <Digit> so why would that not work for your file ? [17:14] <spumco> do you mean channels or streams? i thought your command line was at stream level, not channels [17:15] <spumco> from what i understand files have multiple streams, streams can have multiple channels. is that correct? [17:15] <spumco> my file has 2 streams, video and audio. the audio contains 16 channels. [17:15] <spumco> am i misunderstanding them streams/channels? [17:16] <Digit> i think i am the one not understanding [17:16] <Digit> i hope someone will find a solution for you [17:16] <spumco> no worries, i wasn't sure if i had it wrong [17:16] <spumco> if i work it out, i'll post it here, heh [17:22] <spumco> ok... bin\ffmpeg -i sample-uncompressed-lr-audio.mov -filter_complex "pan=2|FL=c0|FR=c1" stereo.wav [17:23] <spumco> that gives me a stereo file... but it's still silent [17:23] <rexbron> does anyone have any suggestions for dithering or adding noise to an mpeg2 encode? I'm getting some rather noticeable banding when going from 10bit DNxHD to DVD complient mpeg2 [17:27] <rexbron> gradfun states its for playback only [17:33] <spumco> i can't get it to output anything other than silence, heh [17:33] <spumco> bin\ffmpeg -i sample-uncompressed-lr-audio.mov -acodec pcm_s24le -filter_complex "pan=2|FL=1.0*c0|FR=1.0*c1" stereo.wav [17:33] <spumco> that gives me 2 channels of complete silence [17:34] <spumco> the audio codec is the same as the input so the output file is 48KHz [17:34] <spumco> it's just ... silent [17:43] <Digit> try with another codec [17:46] <Digit> you should share that file so other people can try [17:51] <spumco> the file is 1.2 gig, heh [17:51] <Digit> cut it ... [17:51] <spumco> with what? it's a quicktime file [17:52] <spumco> 1920x1080 @ 30 [17:52] <Digit> sigh [17:54] <Digit> cut it with ffmpeg :P [17:54] <spumco> that would be another command line to fight with, heh, i'm already fighting with one [17:54] <spumco> i'm not sure the pan option is working properly [17:54] <spumco> i can specify channels that don't exist and it doesn't complain [17:55] <spumco> this file has 16 audio channels yet i can specify channel 17 with no errors [17:55] <Digit> ffmpeg -ss [start_seconds] / -t [duration_seconds] -i [input_file] [outputfile] [17:55] <Digit> may be your file is just corrupted did you try with another one [17:56] <spumco> i don't think it's corrupted, it came straight off the capture box [17:57] <spumco> it plays ok but i can't hear the audio because it's uncompressed and quicktime won't play that format [17:57] <spumco> i could switch to dnxhd, but i'm pretty sure it's ok [17:57] <Digit> use vlc [17:57] <spumco> vlc won't detect which channels to play [17:58] <spumco> it shows there are 16 48KHz channels, just plays silence [17:58] <spumco> it plays the video just fine, all 7s of it [17:58] <Digit> then how do you know it's not corrupted if you never hear any sound [17:58] <spumco> because i used mplayer to dump the audio [17:59] <spumco> then audacity to load the raw file [17:59] <spumco> the first 2 channels have my audio, the other 14 are empty [17:59] <Digit> so just use virtualdubmod to add the audio to the video [17:59] <Digit> if you have the raw audio [18:00] <spumco> i can get a raw audio file containing 16 channels but i don't know how to crop that into just 2 [18:00] <Digit> and audactity can't do that ? [18:01] <spumco> i could do it in audacity, yes, but that would mean running up a gui app every time i want to convert a file [18:01] <Digit> o_O [18:01] <Digit> well so the real question is why do you get files with 16 channels [18:02] <spumco> you just do, it's not a configurable option [18:02] <Digit> o_O [18:02] <spumco> it's a blackmagic box [18:02] <spumco> shuttle [18:02] <Digit> so weird [18:02] <spumco> yes [18:02] <Digit> can you load you video files in virtualdubmod ? [18:03] <spumco> i tried loading them into virtualdub but that doesn't work [18:03] <spumco> even with the quicktime plugin [18:03] <spumco> what's virtualdubmod? [18:03] <Digit> o_O [18:03] <Digit> it's another "version" [18:03] <Digit> with mkv support and more [18:03] <spumco> the original virtualdub quicktime plugin doesn't understand v210 or avid [18:04] <spumco> is virtualdubmod newer than virtualdub? i shall go look [18:05] <Digit> try it [18:05] <spumco> it's dated 2006! [18:06] <Digit> well the are many other ones [18:08] <Digit> did you try to use ffmpeg to extract only the audio [18:09] <spumco> i did it one channel at a time with pan but i'm not convinced that works [18:09] <spumco> is there a single command to just dump the audio untouched? [18:09] <spumco> to just dump the stream into another file i mean [18:10] <Digit> ffmpeg -i foo -vn -acodec copy output.mp4 [18:11] <spumco> ah, vn = video none i guess, cool [18:11] <spumco> ok, can't put pcm in an mp4 container [18:12] <Digit> o_O [18:12] <spumco> original audio is 24 bit 48 khz pcm [18:12] <LithosLaptop2> put in mkv container [18:12] <spumco> that container doesn't support that format [18:12] <spumco> ok, trying [18:13] <spumco> that gave me a file that plays but is silent [18:13] <spumco> it is pcm 24 le 16 channel 48 khs 24 bit though [18:13] <spumco> is there a way to select channels in vlc? [18:14] <Digit> yes with the menu [18:14] <spumco> the vlc menu has a channel option but it's blank [18:14] <Digit> o_O [18:14] <spumco> jej [18:14] <spumco> heh [18:14] <Digit> your file is a mess :( [18:14] <LithosLaptop2> oes vlc ssay it is 16 channels? [18:14] <Digit> it's a nightmare [18:15] <spumco> yes, codec information says audio, PCM S24 LE (araw), English, Channels:16, Sample rate: 48000 Hz, Bits per sample: 24 [18:15] <spumco> all that looks correct to me [18:16] <spumco> what does "Audio: pcm_s24le ([1][0][0][0] / 0x0001)" mean on the output? [18:16] <spumco> input says "Audio: pcm_s24le (lpcm / 0x6D63706C), 48000 Hz, 16 channels, s32, 18432 kb/s" [18:16] <spumco> output says "Audio: pcm_s24le ([1][0][0][0] / 0x0001), 48000 Hz, 16 channels, 18432 kb/s" [18:17] <spumco> what does all teh "[1][0][0][0]" stuff mean on teh output? [18:18] <LithosLaptop2> ffmpeg -i 14ch.mov -map 0:1 1.wav -map 0:2 2.wav -map 0:3 3.wav -map 0:4 4.wav -map 0:5 5.wav -map 0:6 6.wav -map 0:7 7.wav -map 0:8 8.wav -map 0:9 9.wav -map 0:10 10.wav -map 0:11 11.wav -map 0:12 12.wav -map 0:13 13.wav -map 0:14 14.wav -map 0:15 15.wav -map 0:16 16.wav [18:18] <Digit> o_O [18:18] <LithosLaptop2> try to split all in seperate wavs [18:18] <Digit> lol [18:18] <Digit> good luck :P [18:18] <spumco> heh, i was trying that before but will try it again [18:19] <LithosLaptop2> lol k [18:19] <Digit> there is no sound in your file that's it :P [18:20] <Digit> stop drinking and using ffmpeg :D joke ;P [18:20] <spumco> i thought 0:1 was a stream selector though, not a channel selector? [18:20] <LithosLaptop2> hmm [18:20] <spumco> so channel 0 would be 0.0.0, channel 1 0.0.1 etc. [18:25] <spumco> map just gives me a file with all 16 channels [18:25] <spumco> windows doesn't like it at all [18:25] <spumco> vlc will play it, silently, and show it's 16 channels though [18:33] <Digit> grr cut it a little sample i want to try that shitty video file :P [18:33] <spumco> hehe [18:35] <spumco> <- trying to work out where the audio samples are [18:36] <Digit> lol [18:36] <Digit> those channels are emptyyyyyy :D [18:36] <spumco> heh, it was a very precise video! [18:36] <spumco> if they are, i'm going to kick this pc over and jump on it [18:37] <Digit> ok jump jump :) [18:40] <spumco> ok. i have a 2s chunk of it. 320 meg, heh. [18:40] <spumco> how i get you that chunk, i have no idea [18:42] <spumco> i did this - bin\ffmpeg -ss [start_seconds] -t [duration_seconds] -vcodec copy -acodec copy -i input output [18:43] <spumco> interestingly, it looks different in vlc [18:45] <spumco> the audio codec for stream 1 says "PCM S24 LE (s24l)" in the original and "PCM S24 LE (42ni)" in the chunk [18:51] <spumco> ah, apparently 42ni is what quicktime calls PCM S24 LE [18:51] <spumco> s24l is an alias for the same thing [18:52] <spumco> that's in fourcc.c [18:53] <Digit> share it http://www.filesnack.com/ [18:55] <Bloodhound> resistence is futile [18:55] <Digit> my mother used to say the same thing [18:55] <Bloodhound> :P [18:57] <spumco> you only get 100 meg without paying $14 :( [18:57] <spumco> this file is 320 [18:58] <Digit> grrr [18:59] <spumco> heh, i'm googling for other free shares [18:59] <Digit> http://www.fileconvoy.com/ [19:00] <spumco> filedropper? uploading to that (it was the first one i found) [19:00] <spumco> 5% [19:01] <spumco> is this hobana.freenode.net? (just in case it goes pop when i'm not looking) [19:02] <spumco> 19% [19:06] <spumco> 49% [19:09] <Bloodhound> spumco: hobana is wher *you* are [19:09] <Bloodhound> where* [19:09] <spumco> yes [19:09] <spumco> if i'm in this channel on freenet does it matter which server i'm on? [19:09] <spumco> 75% [19:09] <Bloodhound> not at all [19:09] <spumco> ok [19:09] <sacarasc> It's on freenode. So irc.freenode.net will be able to get here. [19:10] <Bloodhound> unless a netsplit happens... [19:10] <Bloodhound> There are 191 users and 75396 invisible on 26 servers [19:10] <spumco> long time since i've used irc [19:11] <spumco> this file is the worlds most boring file, just fyi [19:12] <spumco> handbrake doesn't understand the audio either sadly :( [19:13] <Digit> o_O [19:13] <Digit> it's an alien file [19:13] <Digit> bewaaaare bewaaaaare [19:14] <spumco> http://www.filedropper.com/showdownload.php/chunk [19:14] <spumco> <- never seen filedropper before, does that work? [19:14] <spumco> it should be a 320 meg mov file [19:14] <Digit> i am broadcasting a movie i have to wait i keep the link [19:15] <Digit> it's not the correct link [19:15] <spumco> nuts, i'll try the one you sent [19:16] <Digit> oh wait it works [19:16] <spumco> \o/ [19:16] <spumco> in vlc you will get video but no audio, i'm guessing it doesn't know what to do with 16 channels [19:16] <spumco> in my quicktime, i get audio but no video - my quicktime won't play uncompressed video [19:19] <spumco> i could probably switch to dnxhd which should, i guess, have the same ffmpeg switches to remap the channels [19:19] <spumco> my dnxhd files also have 16 channels but they play audio/video in quicktime [19:19] <spumco> so they're more comfortable to play with, heh [19:23] <Digit> wow it's not fast [19:23] <spumco> http://www.fileconvoy.com/dfl.php?id=g663c4ba0ec392762999290195942d4bc0fa53… [19:23] <spumco> that's a link to the same file on the file convoy one you mentioned [19:30] <Digit> too late i start with the other one [19:30] <Digit> started [19:34] <Digit> wow man that file got only one channel [19:34] <Digit> wtf [19:36] <spumco> mine? according to quicktimes movie inspector it has 16 48 KHz channels [19:37] <spumco> vlc's codec information also says 16 channels [19:38] <Digit> o_OP [19:38] <Digit> i open it in vlc i can see only one channel [19:39] <spumco> which version? mine is 2.0.6 so i could be out of date [19:40] <Digit> ok 1 flux 16 canaux ... [19:42] <Bloodhound> attempting to stream that is crashing mplayer [19:45] <Bloodhound> [mov,mp4,m4a,3gp,3g2,mj2 @ 0xacea290] moov atom not found [19:52] <spumco> yer, i don't think mplayer liked it on mine [19:52] <spumco> it's v210 i think [19:52] <spumco> <- currently in windows so can't check [19:53] <spumco> back shortly, in linux [19:56] <spumco_> back [19:56] <spumco_> with an _ it seems [20:07] <Digit> how do i open many files with ffmpeg , one after another [20:21] <Digit> nothing there is nothing in the audio of that Sh*|!%? [20:22] <spumco_> i can make it work if i use ffmpeg as a demux/mux [20:22] <spumco_> which i didn't want to do [20:22] <spumco_> i have to go through audacity too to dump the unwanted 14 channels, export them as 24 bit flac stereo [20:23] <spumco_> then use ffmpeg to remux then and convert to 264 [20:23] <spumco_> it works but there's no way to automate the audacity bit [20:23] <spumco_> and it disconnects the video an audio which i didn't want to do [20:23] <spumco_> i used 2 ffmpeg commands, one with -an and one with -vn [20:24] <spumco_> dump teh audio into a 16 channel 24 bit 48 khz pcm raw file [20:24] <spumco_> load that raw into audacity [20:24] <spumco_> dump the 14 channels i don't want then export as 24 bit flac [20:24] <Digit> grrrr it's a nightmare nothing in that audio [20:24] <spumco_> ffmpeg -i vid.mov -i export.flac -vcodec libx264 -map_channel 1.0.0 -map_channel 1.0.1 banana.mp4 [20:25] <spumco_> export the audio with mplayer [20:25] <spumco_> it will dump it as a raw file [20:26] <spumco_> you can import raw then as signed 24 bit PCM, little-endian, 16 channels, 48000 Hz and you'll see the data [20:27] <spumco_> audacity will play it [20:27] <spumco_> you can see the 1st 2 channels with data in them [20:28] <spumco_> you should be able to do "mplayer file.mov -dumpaudio -dumfile out.pcm" [20:28] <spumco_> you should be able to do "mplayer file.mov -dumpaudio -dumpfile out.pcm" [20:30] <spumco_> it won't deal with stripping the channels out when there are 16 [20:30] <spumco_> the only way i can make it work is by dumping the raw audio with mplayer [20:30] <spumco_> ffmpeg won't do it :( [20:32] <Digit> well i try something else later after my broadcasting [20:34] <llogan> spumco_: the danes call it "quality" [20:35] <spumco_> i build it this morning yet it claims to be 1.1.git? [20:35] <spumco_> i can see 1.2.1 listed on the web page [20:35] <spumco_> hmm [20:43] <Digit> forget it ... there is no audio in your hell file ... [20:43] <Digit> 666 [20:44] <spumco_> heh, yes, there is [20:44] <spumco_> export it and you'll see [20:45] <Digit> and not even vlc can extract any audio [20:45] <spumco_> mplayer can dump it [20:45] <spumco_> ffmpeg will dump a 16 channel file that vlc will play as silence [20:46] <Digit> llogan certainly knoes how to do that with ffmpeg if it's possible [20:46] <Digit> but i think that file is really a mess [20:46] <spumco_> i'm nailing command line options for 1.2.1 so i can build that and see if it's different [20:47] <llogan> Digit: only sometimes...also, i don't know what's going on (i just arrived here). [20:47] <spumco_> i have a mov file with 2 streams - video and audio [20:47] <spumco_> the audio is 16 channels [20:48] <spumco_> i want to use ffmpeg to transcode to h264 and take channels 0 and 1 to use as stereo in the output h264 [20:48] <spumco_> all i can get is blank audio [20:48] <spumco_> i'm using map_channel to (try and) do it [20:49] <spumco_> will do, just kicking off a build ^_^ [20:49] <llogan> do you have an input sample? [20:49] <spumco_> let me find the bookmark [20:50] <spumco_> nuts. bookmark is in windows. 2 seconds. [20:51] <spumco_> back shortly. in windows, heh. [20:54] <spumco> sample file should be - http://www.fileconvoy.com/dfl.php?id=g663c4ba0ec392762999290195942d4bc0fa53… [20:55] <Digit> llogan... if you find some audio in that file with ffmpeg... you're a god :D [21:00] <Digit> ffmpeg is still not able to use a playlist ? [21:02] <llogan> Digit: https://ffmpeg.org/trac/ffmpeg/ticket/1199 [21:03] <Digit> : / thanks [21:04] <spumco> command line with output for that chunk file http://pastebin.com/mBjQJZA9 [21:04] <ubitux> a few formats are actually supported [21:04] <ubitux> typically m3u from hls, or asf are supported afaik [21:05] <Digit> probably for linux [21:07] <Digit> for my gui program i use 2 ffmpeg process so i can open another video file without stopping the connection but it's difficult to make it launches the next ffmpeg process just when the current video file is ending [21:20] <spumco> channel mapping without deprecation complaints http://pastebin.com/VsALZLf2 [21:20] <Digit> ? [21:20] <spumco> output file looks ok, audio is still blank though [21:21] <Digit> lmao [21:21] <ubitux> [Parsed_pan_0 @ 00000000023b8140] Pure channel mapping detected: M M [21:21] <Digit> the file from neverland :D [21:21] <ubitux> looks wrong [21:21] <ubitux> the 'M' denotes "mute" [21:21] <ubitux> parsing might have been broken [21:21] <ubitux> spumco: can you open a trac issue? [21:22] <Digit> pure means it drops the channels not selected [21:23] <spumco> i thought pure meant lossless [21:23] <spumco> i was using pan/map_channels as a selector to drop the 14 channels i didn't want [21:23] <Digit> it's lossless because it just needs to drop some channels [21:23] <spumco> i didn't know the "M" meant mute, where does it say that? [21:24] <spumco> ubitux, the only sample file i have it pretty large [21:24] <Digit> i think it's middle not mute [21:24] <llogan> 321M for 2 seconds [21:24] <ubitux> Digit: no it's mute [21:25] <ubitux> spumco: try to extract only the audio, and keep 4-5 seconds [21:25] <spumco> llogan, it's uncompressed quicktime 1920x1080@30/16 ch 48KHz PCM le 24 bit [21:25] <spumco> the video goes across ok [21:26] <spumco> <- looking at web copy of af_pan source ^_^ [21:26] <ubitux> spumco: ffmpeg -i chunk.mov -vn -c:a copy -t 20 out.flac [21:26] <ubitux> or maybe well out.wav [21:26] <ubitux> if you can reproduce the issue with the wav, then please open an issue [21:27] <spumco> that line gives me unsupported codec [21:27] <spumco> at least on this o/s it does, heh [21:31] <spumco> audio alone is 4.39 meg [21:34] <spumco> will build latest tomorrow then try again before submitting anything [21:34] <spumco> thank you o/ [21:38] <Digit> do you know what 's the correct rtmp url to use ? i got the ustream url for my account and key but when i use the key after the url it doesn't work [21:44] <rafael2k> people, do you know any patch to enable xHE-AAC encoding/decoding for ffmpeg? [21:44] <TomHoss> Anyone know if it's possible to stream a flv file from my server, when I have the url of the file on an external website? [21:44] <rafael2k> linking to the ISO reference code [21:44] <TomHoss> I'm wondering if I can just stream it through and avoid downloading the whole thing first [21:44] <TomHoss> I can do it with a webm file, but not flv [21:44] <Digit> o_O [22:10] <llogan> rafael2k: do you have a sample? [22:16] <Digit> woooowoooow i found how to stream to ustream :) [22:18] <Digit> ffmpeg rocks on ustream B) [22:19] <llogan> what does your url look like now (excluding your actual account key)? [22:21] <Digit> i needed to add flashver=FMLE/3.0\20(compatible;\20FMSc/1.0) [22:22] <Digit> it's like this -f flv "rtmp://UstreamURL/UstreamKey flashver=FMLE/3.0\20(compatible;\20FMSc/1.0)" [22:22] <Digit> B) rocks ! :) [22:28] <Digit> i start to know some little things about ffmpeg :D [22:29] <Digit> finally what was the correct command for the guy with 16 audio channels and wanting to keep 2 ? does it work now ? [00:00] --- Sun May 19 2013

1 0

ffmpeg-devel.log.20130517
by burek 18 May '13

18 May '13

[00:11] <ubitux> trac is having a hard time tonight [00:11] <ubitux> it is extremely slow.. [00:19] <Daemon404> ubitux, it's always been quite slow for me [00:19] <Daemon404> michael could never reproduce tho [00:19] <Daemon404> iirc [00:20] <michaelni> ubitux, are you using trac.ffmpeg.org or ffmpeg.org/trac ? and how to reproduce ? [00:21] <ubitux> it's ok now [00:22] <ubitux> seems i was on ffmpeg.org/trac [00:22] <ubitux> it was just for a few minutes [00:22] <ubitux> it's fast now :) [00:32] <ubitux> burek: could you fix the README.html in the {32,64}bit directories of your static builds? [00:32] <ubitux> it seems to confuse users [00:32] <ubitux> (and i'm not sure empty readmes make much sense) [00:36] <cone-801> ffmpeg.git 03Michael Niedermayer 07master:ab074ab9fef8: avcodec/pthread: fix ffv1 with 2 threads [00:45] <cone-801> ffmpeg.git 03James Almer 07master:54372848f327: tools/ffhash: Use O_BINARY when available [00:47] <cone-801> ffmpeg.git 03Stefano Sabatini 07master:0aa013fa8a5f: lavfi: factorize drawgrid and drawbox code [00:47] <cone-801> ffmpeg.git 03Stefano Sabatini 07master:cb0b0266e5c9: lavfi/drawbox: restore verbal form for drawbox options descriptions [03:33] <cone-801> ffmpeg.git 03Carl Eugen Hoyos 07master:17989dcf540c: configure: Autodetect VA-API [03:52] <highgod> Hi, I want to ask a question about audio, I found that ffmpeg out put a frame after decoding audio, and I want to save the content of it, what format should I save, wav or something? how can I play it, thanks. [07:36] <burek> ubitux, try to View Source on those README.html files and you'll get the answer why do I keep them :) [07:40] <burek> but I'll try to find a better way to do that css styling [07:40] <burek> not to confuse people [09:33] <burek> at http://www.ffmpeg.org/ffmpeg-filters.html#fps [09:33] <burek> there is a 2nd example: fps=fps=film:round=near [09:34] <burek> "film" is never listed anywhere in this document, so it might be a good idea to insert the list of such abbreviations? [09:56] <saste> burek, ffmpeg-utils(1) [09:58] <burek> saste, you mean man ffmpeg-utils [09:58] <burek> or something else? [10:00] <burek> oh, you mean http://www.ffmpeg.org/ffmpeg-utils.html ok [10:00] <burek> thanks [10:01] <burek> would it be convenient that we add just a link to the http://www.ffmpeg.org/ffmpeg-utils.html#Video-rate [10:01] <burek> not to duplicate the content [10:04] <highgod> Hi, all ,I want to ask a question, I saw that the audio is decoded in a avframe, is it a pcm format? [10:18] <cone-86> ffmpeg.git 03Jeremy Hinegardner 07master:28e6b7b9b4e0: Turn off usage of base_data_offset in tfhd for PIFF files [10:26] <burek> when i think about it, it's better not to add any link to abbreviations for fps, since the structure of the documentation can be changed at any time and the link might break.. [10:26] <BuxiNess> Morning [10:27] <burek> anyway, some reference to those abbreviations is needed in the fps filter docs, if we're going to use such an example, so that people can see the list of all possible abbrev. [10:34] <cone-86> ffmpeg.git 03Sebastian Sandberg 07master:1fb013a57c6d: vc1dec: Add support for interlaced B-frames [10:34] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:c453da33a3de: Merge commit '1fb013a57c6d98d3f425506eb25f32b2cf7ddc6f' [10:37] <burek> one more bug, http://www.ffmpeg.org/ffmpeg.html describes -loop_input and -loop_output, referencing -loop (because first two are deprecated), but there is no entry for -loop ? [10:39] <michaelni> morning BuxiNess [10:41] <BuxiNess> morning [10:46] <michaelni> BuxiNess, as you are maintainer of jpeg2000 you should have op here [10:48] <BuxiNess> michaelni, ok [10:49] <cone-86> ffmpeg.git 03Jeremy Hinegardner 07master:40bbbce7e8ad: movenc: Don't add base_data_offset in tfhd for PIFF files [10:49] <cone-86> ffmpeg.git 03Kostya Shishkov 07master:1232a1647ab2: Apple Intermediate Codec decoder [10:49] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:5bc892cb1722: Merge commit '1232a1647ab27e024a3baf4d01d40c8d08d6ced9' [10:59] <cone-86> ffmpeg.git 03Anton Khirnov 07master:b06848f4decc: vf_lut: make config_props work properly when called multiple times. [10:59] <cone-86> ffmpeg.git 03Anton Khirnov 07master:87c31cfd7a2c: vf_gradfun: make config_props work properly when called multiple times. [10:59] <cone-86> ffmpeg.git 03Anton Khirnov 07master:d371c3c2e283: vf_frei0r: make config_props work properly when called multiple times. [10:59] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:d5e5e73c52d3: Merge commit 'd371c3c2e2830d9783465ecfe1ab7d93351083b7' [11:11] <cone-86> ffmpeg.git 03Anton Khirnov 07master:7727be79d1ee: vf_fps: move initializing pts from config_props to init. [11:11] <cone-86> ffmpeg.git 03Anton Khirnov 07master:90d9a2a04c3f: vf_drawtext: do not reset the frame number in config_input. [11:11] <cone-86> ffmpeg.git 03Anton Khirnov 07master:4c205f42c86c: vf_drawbox: make config_props work properly when called multiple times. [11:11] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:c7078f4907dc: Merge commit '4c205f42c86ccefa093c59434669af34ad14a52b' [11:40] <cone-86> ffmpeg.git 03Anton Khirnov 07master:d5c66d9c5612: vf_setdar: make config_props work properly when called multiple times. [11:40] <cone-86> ffmpeg.git 03Anton Khirnov 07master:6592cd22a230: vf_crop: make config_props work properly when called multiple times. [11:40] <cone-86> ffmpeg.git 03Anton Khirnov 07master:3fb29588a27a: vf_drawtext: don't leak the expressions. [11:40] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:71fcb6072ea6: Merge commit '3fb29588a27a711132106b924e27b53789a58dcb' [11:47] <cone-86> ffmpeg.git 03Anton Khirnov 07master:3ba35a346cd2: vf_hqdn3d: make config_props work properly when called multiple times. [11:47] <cone-86> ffmpeg.git 03Anton Khirnov 07master:16a645adeb75: vf_pixdesctest: make config_props work properly when called multiple times. [11:47] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:1e9a050bc020: Merge commit '16a645adeb758207346a4bbf66766f02734c461e' [11:53] <ubitux> https://www.youtube.com/watch?v=K6JshvblIcM [11:55] <cone-86> ffmpeg.git 03Anton Khirnov 07master:30e256430eb8: lavc/pthread: remove obsolete checks [11:55] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:36d349034e19: Merge commit '30e256430eb88c6d4c382581b89bca171d79fbc0' [12:01] <cone-86> ffmpeg.git 03Anton Khirnov 07master:db8e336e5fc1: pthread: reindent [12:01] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:1a0057dc881c: Merge commit 'db8e336e5fc18c41cb4e9bab1960349e9800d546' [12:08] <burek> ubitux +1 [12:18] <cone-86> ffmpeg.git 03Jindrich Makovicka 07master:259af1b92370: mpegvideo: allocate sufficiently large scratch buffer for interlaced vid [12:18] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:1ec578c010e7: Merge remote-tracking branch 'qatar/master' [12:49] <ubitux> i wonder if those config_props make that much sense in all cases [12:49] <ubitux> from what i can tell, a lot of filters could just have a flag to say they are "uninit" compliant [12:49] <ubitux> which means uninit() could be called before config_props() [12:50] <ubitux> well, let's just wait to see what anton will come up with [12:58] Action: durandal_1707 *powerpoint [13:13] <michaelni> smarter, how is your decoder going ? can something already be merged ? [13:14] <durandal_1707> michaelni: what's about alpha in rgba64 output from non-alpha source, is that initialization really needed? [13:14] <smarter> michaelni: not yet, but it's starting to get into shape [13:15] <smarter> michaelni: beside various cleanups, I hope to be able to merge the parallel decoding stuff that is being worked on in https://github.com/OpenHEVC/openHEVC/tree/hm10.0 [13:17] <michaelni> smarter, ok [13:18] <michaelni> also dont hesitate to ping me if theres something to merge into ffmpeg or something to review [13:18] <michaelni> or if you have questions about any code from h264*.c [13:19] <smarter> sure, thanks [13:19] <michaelni> durandal_1707, the alpha in rgba64 that is output should be initilaized to "opaque" [13:19] <smarter> I looked into reusing some of h264_parser.c but it seems pretty specific to h264 [13:20] <durandal_1707> michaelni: yes, but i dont see similar code for 8bit case [13:21] <michaelni> smarter, yes, the startcode search code could be reused i guess unless its already done [13:21] <michaelni> durandal_1707, do we output proper opaque alpha for 8bit ? [13:22] <smarter> I think ff_h264_find_frame_end looks for things like end of slices that are different in h264 and hevc [13:26] <michaelni> it parses the first mb index and nal type, id have to cross check h264 & 5 to know how much that differs [13:29] <michaelni> also it contains code to parse the length values that are used in h264 in mp4 but not in raw h264 [13:32] <durandal_1707> michaelni: you incorrectly merge doc/APIchanges [13:33] <cone-86> ffmpeg.git 03Paul B Mahol 07master:3a751eab7860: lavu/opt: add AV_OPT_TYPE_COLOR [13:33] <cone-86> ffmpeg.git 03Paul B Mahol 07master:05b7560d39ab: lavfi/color: switch to AV_OPT_TYPE_COLOR [13:33] <cone-86> ffmpeg.git 03Paul B Mahol 07master:83e892233c87: lavfi/life: switch to AV_OPT_TYPE_COLOR [13:33] <cone-86> ffmpeg.git 03Paul B Mahol 07master:1d2ecf9609a7: lavfi/drawtext: switch to AV_OPT_TYPE_COLOR [13:33] <cone-86> ffmpeg.git 03Paul B Mahol 07master:85e8a1169cfb: lavfi/pad: switch to AV_OPT_TYPE_COLOR [13:41] <michaelni> durandal_1707, please elaborate, which commit did i mess up in which way ? [13:43] <durandal_1707> the .0 changes from libav are in wrong place [13:44] <michaelni> you mean the whole is not in chronological order ? [13:49] <michaelni> I admit i did not pay much attention to the order, ill try to pay more attention to it from now [13:49] <michaelni> about the existing entries, someone should fix their order [13:50] <michaelni> i dont volunteer though [13:55] <Compn> Apple Intermediate Codec decoder [13:55] <Compn> gasp [13:56] <Compn> :) [13:56] <Compn> another codec down! [14:59] <durandal_1707> ubitux: is #1907 related to subtitles? [15:02] <ubitux> no idea [15:02] <ubitux> possibly [15:02] <ubitux> i have no idea how to reproduce though [15:03] <durandal_1707> seek [15:05] <ubitux> i don't see any sample [15:06] <ubitux> is this detected by valgrind btw? [15:09] <durandal_1707> address sanitizer [15:09] <ubitux> yeah but i don't have it, so i'm wondering if i can reproduce with valgrind [15:20] <cone-86> ffmpeg.git 03Vignesh Venkatasubramanian 07master:8ffcc826b3f3: Adding support for muxing VP8 Alpha files [15:20] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:2fb193b1e742: vformat/matroskaenc: factorize alpha code in mkv_write_block() [15:20] <cone-86> ffmpeg.git 03Vignesh Venkatasubramanian 07master:98c292a7a8f2: Adding support for encoding VP8 Alpha [15:20] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:ebe8c7fe5201: vc1dec: restructure interpmvp code [15:34] <Compn> pew pew pew [15:34] Action: Compn shoots at cehoyos [16:20] <durandal_1707> ubitux: can i push to files you are maintainer(libmodplug) without specific approval? [16:21] <ubitux> i guess :) [16:21] <ubitux> you can just show me the commit here if you want a quick review [16:21] <ubitux> you want to adjust probing? [16:22] <ubitux> if you wish to remove 'mod' as an extension i don't think that's a good idea [16:22] <ubitux> since that's a very common extension [16:22] <ubitux> for midi-like files [16:25] <durandal_1707> ubitux: i did just s/int sz/int64 sz/ [16:25] <durandal_1707> for >2 gb mod files :))))) [16:26] <ubitux> not a good idea [16:26] <ubitux> because there is a cache [16:26] <ubitux> there is a design issue in libmodplug [16:26] <ubitux> which forces you to store the whole file in memory iirc [16:27] <durandal_1707> avio_size returns int64, which in code gets into int [16:27] <ubitux> ah, well ok [16:27] <ubitux> feel free to fix that [16:32] <durandal_1707> Compn: software to show at people? [16:33] <durandal_1707> *shot lol [16:34] <Compn> durandal_1707 : haha , well you know ... dont want to work on software that can be used against you [16:34] <Compn> :P [16:34] <Compn> skynet, etc [16:35] <durandal_1707> we really need such filter [16:35] <Compn> what? facial recognition ? [16:35] <durandal_1707> recognition of interesting stuff.... [16:35] <durandal_1707> *detection.... [16:35] <Compn> ah [16:43] <durandal_1707> hmm so file is probed as mod(score 50) instead mpeg(score 52) just because libmodplug have higher priority than mpeg for some reason [16:43] <ubitux> 'o' is before 'p'? :p [16:44] <durandal_1707> but i thougt that all are probed and the one with highest score wins [17:00] <Compn> 2gb mod files? what? [17:39] <durandal_1707> no more comments on ftp patch? [18:13] <cone-86> ffmpeg.git 03Paul B Mahol 07master:d26f637b020e: libmodplug: fix sz type [18:44] <durandal_1707> anyone have java installed? [18:47] <durandal_1707> ubitux: what's status of vignette filter [18:47] <ubitux> i need to add dithering [18:48] <ubitux> and eventually add other formulas since michaelni seems to be expecting something different [18:49] <durandal_1707> sure that "vignette" name would be correct term in such case? [18:49] <ubitux> i could rename to lens [18:49] <ubitux> but that's a limited use of "lens" [18:57] <cone-86> ffmpeg.git 03Paul B Mahol 07master:47ac344970f1: id3v2enc: chapter support [19:35] <j-b> BBB: do you know people working on libyuv? [19:35] <JEEB> did you notice the mingw compilation failure? [19:39] <BBB-work> j-b, yes [19:39] <BBB-work> j-b, any specific message I should convey? [19:41] <j-b> BBB-work: is it good or as bad as it looks like ? :D [19:43] <BBB-work> j-b, ? [19:43] <BBB-work> j-b, you'll have to be more specific [19:43] <BBB-work> I can interpre that in many ways [19:43] <j-b> BBB-work: nothing specific, I fell on it, and wondered how good it was to integrate in VLC or not [19:44] <j-b> BBB-work: (I was looking for fast i420->nv12 conversion) [19:45] <nevcairiel> i420 -> nv12 is so simple that even thinking about getting a lib for that is just insane :P [19:46] <j-b> nevcairiel: read better what I said. [19:47] <nevcairiel> its also simpel to make it fast [19:47] <j-b> with pump* yes [19:47] <nevcairiel> find someone with just a tiny bit of SIMD experience on your target platforms, really not much shuffling required [19:47] <j-b> but I was looking for a simple good example to show to my students [19:48] <j-b> and I fell on libyuv [19:51] <j-b> because even if "simple" there are not many good examples on the net. [19:52] <nevcairiel> this is true [19:52] <j-b> and more examples of the opposite [19:52] <j-b> swscale code scares me [19:52] <gnafu> Hehe. [19:59] <BBB-work> j-b, it's fast. quality-wise it doesn't always do the very-best (read: slower) thing, but it's fast and chrome uses it [20:01] <Compn> hardware acceleration in a web browser :\ [20:01] <Compn> j-b : you should show them swscale. either they learn how not to do it, or one of them understands it and can come help us with it :D [20:02] <j-b> BBB-work: ok [20:02] <j-b> :) [20:02] <Compn> people creating video filters to do colorspace conversions instead of adding it to swscale :\ [20:02] <Compn> (re vf xyz ) [20:05] <BBB-work> it's not like xyz is sweet [20:48] <nevcairiel> Compn: swscale can convert from xyz to rgb now [21:11] <ubitux> so windows build are stalled? [21:11] <ubitux> (zeranoe) [21:11] <ubitux> because of the windows bug? [21:12] <michaelni> didnt i read the bug was fixed ... [21:13] <ubitux> no idea [21:13] <ubitux> the builds on zeranoe are not up-to-date [21:13] <ubitux> and it happens that a user need a feature recently added [21:14] <nevcairiel> the bug in mse was indeed fixed [21:14] <nevcairiel> i could produce a static build without external libs if anyone needs it desperately [21:15] <michaelni> ubitux, did you or someone else contact zeranoe and inform him that the bug has been fixed ? if not someone shoukd [21:15] <michaelni> shouLd [21:16] <ubitux> i don't plan to do [21:17] <ubitux> heh seems the guy get it working anyway [21:17] <ubitux> derp [21:17] <ubitux> the build is up-to-date [21:18] <ubitux> just a brain fart, forget all i said. [21:19] <cone-86> ffmpeg.git 03Reimar Döffinger 07master:24c65eb29f05: md5: avoid unnecessary copying. [21:21] <TheRyuu> 21:14 <+nevcairiel> the bug in mse was indeed fixed <---what did they break this time? [21:21] <nevcairiel> it crashed when scanning swscale [21:21] <ubitux> after xyz addition [21:21] <TheRyuu> el oh el [21:21] <JEEB> yeh [21:21] <JEEB> it was a funny thing [21:22] <JEEB> I think I even once got worried what on earth happened when MSE just told me it had gotten disarmed [21:22] <TheRyuu> I don't think they 'fixed' the whole concurrent write thing yet though [21:22] <ubitux> if (TheApplicationName == "swscale") return 0; [21:22] <TheRyuu> although that's not really a problem if it's a gcc build [21:23] <TheRyuu> mine doesn't like the example program when compiling zlib [21:23] <TheRyuu> herp derp antivirus [21:25] <ubitux> so the md5 optim just added is supposed to make fate faster? [21:26] <ubitux> or the switch to crc/murmur3 was done instead already? [21:30] <michaelni> its all improving in small steps, faster md5 is good for more than fate [21:30] <michaelni> fate wasnt switched yet to something else [21:32] <ubitux> ok [21:36] <Compn> nevcairiel : i know . but i'm saying the person who wrote the filter couldnt figure out how to add xyz to swscale [21:36] <Compn> and we should fix that [21:36] <Compn> if possible [22:16] <durandal_1707> "bugs only encountered by drones" ... [22:17] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:6c593f1b671b: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/0.10:1c655c1419af: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/0.11:f7dbe180a02f: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/0.9:b22cc0c15de7: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/1.0:663b22325891: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/1.1:f544553c2993: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/1.2:fd60aeb55b12: avidec: dont randomly skip packets for offseting the index [23:03] <cone-86> ffmpeg.git 03Carl Eugen Hoyos 07release/1.2:af589dd5e9f8: configure: Autodetect VA-API [23:28] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/1.0:14634f5b698d: ffmpeg: free threads on error conditions. [23:28] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/1.1:85277ff9364c: ffmpeg: free threads on error conditions. [23:28] <cone-86> ffmpeg.git 03Michael Niedermayer 07release/1.2:ffd28de38863: ffmpeg: free threads on error conditions. [00:00] --- Sat May 18 2013

1 0

ffmpeg.log.20130517
by burek 18 May '13

18 May '13

[00:05] <relevante> ubitux: should I paste all of those logs into the ticket directly, or is using a paste service preferred? [00:06] <relevante> the complete log is over 14k lines [00:06] <ubitux> relevante: no it's ok [00:06] <ubitux> i've added the relevant information, i believe [00:07] <relevante> cehoyos asked again for the "complete, uncut console output" [00:08] <ubitux> relevante: i've pasted what's necessary [00:08] <ubitux> relevante: do you have other decoder we can compare against? [00:10] <relevante> is a player ok? [00:10] <ubitux> you can point out anything able to play the file [00:11] <relevante> ok sure [00:11] <relevante> having trouble accessing trace now [00:11] <relevante> trac* [00:12] <ubitux> yeah, same here :) [00:17] <relevante> ok great, it's back [00:22] <mgeary> hi folks [00:23] <mgeary> i'm trying to convert a fly audio file to ogg. I'm getting the error: "Unable to set CBR to 22050: not supported" [00:23] <mgeary> what do i need to change? [00:25] <mgeary> ubitux: http://pastebin.com/BLanLRTN [00:25] <ubitux> wow this is old :) [00:25] <ubitux> it works fine here with a more recent version [00:26] <mgeary> interesting [00:26] <mgeary> iirc, it was no small trick to get this compiled and configured on this machine [00:26] <mgeary> so i'm reluctant to go through that again :S [00:26] <ubitux> you can try another audio bitrate anyway [00:26] <ubitux> like 44100 [00:26] <ubitux> (-f ogg should not be necessary btw) [00:27] <mgeary> every other rate i try fails in the same way [00:27] <mgeary> can i just keep it CBR? [00:27] <mgeary> if i don't specify a bitrate, it works, but my file ends up being huge [00:28] <ubitux> it's hard to tell; 0.10 is too old i have no idea how it behaves [00:29] <mgeary> this machine is Ubuntu 10.04, is there an apt-get repo i could point to to get a newer version? [00:29] <mgeary> i had to compile this version by hand, iirc, looping in a bunch of plugins [00:30] <mgeary> yeah, but those don't have mp3 support compiled in, do they? [00:30] <ubitux> i see "--enable-libmp3lame" in the configure options [00:31] <mgeary> hmm [00:31] <mgeary> where can i see those options? [00:31] <mgeary> the readme seems empty [00:31] <mgeary> oh [00:31] <mgeary> i'm down too deep. sorry [00:31] <mgeary> was in the 64-bit dir [00:35] <mgeary> hmm. no --enable-libfaac. Does that mean i won't be able to read AAC files? [00:35] <ubitux> no, you just won't be able to encode them with libfaac [00:35] <mgeary> don't need to encode. just decode [00:35] <ubitux> (but libvo-aacenc is available in that build) [00:35] <ubitux> decode is native [00:35] <ubitux> just like mp3 [00:36] <mgeary> hmm. trying with the latest static build i now get: [libvorbis @ 0x2792820] encoder setup failed [00:36] <mgeary> pastebinning [00:37] <mgeary> http://pastebin.com/M8dkmMTd :S [00:38] <ubitux> oh i'm stupid [00:38] <ubitux> i didn't realized you were using audio *bitrate* [00:38] <ubitux> you are looking to -ar, not -ab [00:38] <ubitux> s/to/for/ [00:39] <mgeary> huh [00:39] <mgeary> ah, sampling rate versus data rate, yes? [00:39] <mgeary> duh [00:40] <mgeary> that actually might explain why a mp3 file i produced isn't playing back as expected in IE [00:43] <ubitux> relevante: as a workaround, you can try to revert the hash cehoyos pointed out [00:44] <relevante> great, will try it now [00:53] <relevante> ubitux: i'm getting this build error: [00:53] <relevante> Undefined symbols for architecture x86_64: [00:53] <relevante> "_MM_FIX_0_541196100", referenced from: [00:53] <relevante> _filter in libavfilter.a(vf_fspp.o) [00:53] <relevante> "_MM_FIX_0_707106781", referenced from: [00:53] <relevante> _filter in libavfilter.a(vf_fspp.o) [00:53] <relevante> ld: symbol(s) not found for architecture x86_64 [00:53] <ubitux> relevante: did you check out the revision or just revert the commit? [00:53] <relevante> sounds like i need to force gcc instead of letting it use llvm. do you know off the top of your head how i do that [00:53] <relevante> i checked out [00:54] <ubitux> stay on master [00:54] <ubitux> and revert the commit [00:54] <relevante> oh ok [00:54] <relevante> standby :) [00:54] <ubitux> or --disable-filter=mp should be enough to fix that build [00:54] <relevante> on master? [00:54] <relevante> with that flag [00:57] <ubitux> on master you don't need it [00:57] <ubitux> it's fixed afaik [00:57] <ubitux> (or maybe not) [00:57] <relevante> ok cool [00:57] <relevante> thanks [00:58] <ubitux> iirc it's a llvm failure [00:58] <ubitux> i wonder now if we workarounded it [00:59] <relevante> hmm, i have conflicts on the revert [01:00] <relevante> https://gist.github.com/relevante/4dbcc6a370e1068ec86f [01:00] <relevante> https://gist.github.com/relevante/7ad4126149567a139a31 [01:01] <relevante> i guess i should take the old one everywhere? [01:05] <relevante> oops, guess not. that didn't work [01:11] <roboman2444> whats the best way to use ffserver. ffmpeg, and v4l2 to stream a basic webcam? [01:12] <roboman2444> i dont want to waste processing power converting the mjpeg output of the webcam to something else [01:15] <roboman2444> If i dont set things like bitrate or framerate, will ffmpeg just pass the video data through and not try to encode it at all? [01:21] <relevante> ubitux: I couldn't get the reverted build to work. i also tried checking out the old one (c9c7a77) and building it with the flag you suggested, which built, but still doesn't convert the file properly. I also tried the previous commit with the same result. [01:21] <relevante> looks like they boosted the priority of the ticket though, which is good :) [01:55] <Pcmad> it seems i can't use windows path to my file when i want to use the overlay filter with a image file... any patch or any other version i could use to fix it ? [02:13] <Magicking> Pcmad: As a workaround can't you execute ffmpeg with the current working directory defined as the directory of your image ? [02:13] <Magicking> I know, it's dirty [02:20] <Pcmad> i can use the same directory as the input but it's a problem to do like that [02:21] <Pcmad> and when i ffmpeg the result with rtmp i get no overlayed image [02:21] <Pcmad> to rtmp [02:22] <Pcmad> the is no way to fix that window's path problem ? [03:38] <roboman2444> bleh [03:38] <roboman2444> i need a ffmpeg that runs on a 486 [03:38] <roboman2444> dont ask [03:43] <roboman2444> flags : fpu tsc cx8 [03:43] <mapreduce> I'm using valgrind on a program I'm writing and it's complaining about stuff within av_calloc [03:44] <mapreduce> Is it normal for ffmpeg's code to give valgrind warnings? [03:44] <mapreduce> libav* [03:53] <highgod> Hi, I want to ask a question about audio, I found that ffmpeg out put a frame after decoding audio, and I want to save the content of it, what format should I save, wav or something? how can I play it, thanks. [06:23] <Guest90973> HHello all [06:23] <Guest90973> for some reason my nickname changed back =( [06:23] <Guest90973> can i possibly bother someone for some help? [06:26] <Guest90973> anyoen? [09:49] <xlinkz0> can i apply a filter for a specific part of the video and copy the rest? [09:59] <saste> xlinkz0, check timeline [10:01] <burek> xlinkz0, yes you can [10:07] <xlinkz0> well i think the fade filter already does that [10:07] <xlinkz0> but i basically want to c:v copy after it is finished [10:08] <xlinkz0> is it possible not to transcode the whole video just for the few seconds of fade? [10:10] <highgod> Hi, all ,I want to ask a question, I saw that the audio is decoded in a avframe, is it a pcm format? [10:12] <burek> xlinkz0, no, I thinkt it's not [10:13] <burek> using -vf assumes re-encoding [10:14] <burek> xlinkz0, try using some advanced video editor to cut your original video on keyframes, that surround your interval where you want your filter applied, use the filter only on that chunk and try to merge it back into original video [10:14] <xlinkz0> so i have to make 3 cuts.. [10:14] <burek> this might be easier if you use HLS, since you'll be working with chunks anyway [10:14] <xlinkz0> hls? [10:14] <burek> google it [10:15] <xlinkz0> that's a vastly used acronym.. [10:15] <burek> highgod, it might be, but it could also be some internal format that makes ffmpeg work more fast, etc [10:15] <burek> highgod, why do you ask anyway? [10:16] <burek> oh ok [10:17] <xlinkz0> oh i'm not using that, i need to cut parts from various mp4's and concat them together but with certain transition effects [10:17] <xlinkz0> they already have a keyframe each second so i can cut fast [10:28] <highgod> burek:we decode it out and pass it to the next module to use, so we don't konw what format it is and we want to test whether the result we got is correct [10:29] <burek> highgod, did you take a look at the source code? [10:29] <highgod> yes [10:29] <highgod> audio docode [10:29] <highgod> burek:yes, but not firmilar [10:29] <burek> well, did you get the idea in what format is it ? [10:29] <burek> which* [10:31] <highgod> no, just copy the data to a file [10:54] <xlinkz0> is there any transition filter that acts something like zooming? [10:55] <xlinkz0> like scale a frame to 10% put it at x,y then slowly scale it bigger and bigger until it fills the screen [10:55] <xlinkz0> i mean scale consecutive frames bigger and bigger [10:58] <highgod> burek:the fromat is AV_SAMPLE_FMT_FLTP [11:04] <highgod> buerk:is there any player to player this format audio? [11:08] <highgod> Can I use the swscale to transfer it to pcm? [11:11] <burek> xlinkz0 ^ [11:11] <burek> highgod, my point is that you shouldn't do it the way you are describing it [11:12] <burek> ffmpeg uses (or can use, in the future) the internal format suitable for faster processing of audio [11:12] <csteve> hi everyone, it's the first time i use irc, and i have two question about ffmpeg dev...anyone can help me? [11:12] <burek> so, you shouldn't rely on its structure for your ongoing projects, rather you should encode the audio into some standard audio format [11:13] <retard> csteve: channels have topics, you might want to look at the one for this channel [11:13] <burek> hi csteve, just feel free to ask your questions [11:15] <csteve> thank, so i'm implementing ffmpeg for mpeg2 xdcamhd422 encoding, works already well, but i would like to be clear with two encoding options [11:16] <highgod> burek:but I don't konw whether the data we get is correct, do you konw which player can play this format? [11:16] <burek> highgod, let me repeat, this time more clearly: no player, except ffmpeg, should ever be able to play any ffmpeg-internal data format [11:17] <bomzaiya> hi I am Bom from Thailand [11:17] <burek> that's kinda the point of internal formats [11:17] <burek> highgod, try using latest ffmpeg to convert your input audio into pcm output (wav) and see if the output is correct [11:18] <burek> that would pretty much mean that internal representation of that same audio is correct too [11:18] <highgod> burek:if I transcode it to AV_SAMPLE_FMT_S16, then it can use pcm player to pcm [11:18] <highgod> yes [11:18] <bomzaiya> Any one has experience on opening transparent video on android? [11:19] <highgod> burek:but how can I imput the file? it has no format,wav or .... [11:19] <bomzaiya> Or what are commands I could tell ffmpeg to give me respond back to java? [11:23] <csteve> First question: how can i set the "qscale" parameter of the codec, and secondly how can i set the color matrix coefficient? [11:45] <burek> highgod: ffmpeg -i input.mp4 output.wav [11:46] <bomzaiya> hi [11:46] <burek> bomzaiya, you might try: ffmpeg -i INPUT -c copy -c:v libx264 - [11:47] <burek> "-" at the end means that ffmpeg should redirect the output stream to the stdout [11:47] <burek> which you might capture in your java app [11:47] <bomzaiya> oh [11:47] <bomzaiya> interesting. [11:48] <burek> csteve, did you try reading the ffmpeg documentation first? [11:49] <bomzaiya> I read. [11:49] <bomzaiya> But I might miss something like that. [11:49] <bomzaiya> actually, I only saw those about conversion. [11:51] <burek> bomzaiya, try here: http://www.ffmpeg.org/ffmpeg-all.html [11:51] <bomzaiya> ok thx [12:18] <deviant695> hi all [12:18] <deviant695> i am doing hls and using the segmenter but i am having a problem with the adaptive streaming [12:19] <deviant695> the video starts form the begining when i change bitrate trough jw player [12:19] <deviant695> i have tried many variants but nothing so far [12:19] <deviant695> can anyone sugest something? [13:52] <deviant695> welli guess its useless when nobody is active here [14:08] <xlinkz0> say i input a lot of files with -i 1.mp4 -i 2.mp4 etc, how do i reference these in filtergraphs? [14:09] <deviant695> do not bother nobody will reply [14:12] <durandal_1707> xlinkz0: read documentation, its mentioned there [14:13] <xlinkz0> L1, L2 ? [14:16] <durandal_1707> names can be anything [14:19] <durandal_1707> with -filter_complex, and without any mapping inputs are named by default like this: [0], [1].... [14:32] <xlinkz0> thanks [14:50] <t4nk684> i can use the overlay filter with an image file and output to a flv file but when i try to output to rtmp i get no overlay.... any idea ? [14:58] <t4nk684> it seems i found the problem but i don't know how fix it, the overlay filter works when i remove the scale filter... why ? do we need to write both filters in the same vf= ? [15:05] <t4nk684> i finally found the solution :) i had to seperate the filters with commas [15:06] <t4nk684> but the window's path for the image doesn't work it's a bug we can't use a : to specify a drive [15:06] <t4nk684> it's very boring [15:07] <t4nk684> do you know a patch for that problem ? [15:47] <Mickey_> hi ! i overlay my animated image over a video but it uses only the first image ?! [15:51] <Mickey_> http://pastebin.com/J5zDPbRM [15:52] <durandal_1707> that is incomplete and cut output, only way i can help you is if you provide exactly what was requested [15:52] <Mickey_> incomplete ??? [15:53] <Mickey_> not it's not ... [15:53] <Mickey_> no it's complete o_O [15:54] <ubitux> "and the COMPLETE console output" [15:54] <Mickey_> it works but the image is not animated [15:56] <ubitux> my guess is that you are using a too old ffmpeg, so without animated gif support [15:56] <ubitux> but no one can tell since your output is incomplete [15:56] <Mickey_> it's the last version i downloaded it 3 days ago [15:56] <ubitux> what is "last version"? [15:56] <durandal_1707> and wrom where is it downloaded? [15:56] <durandal_1707> *from [15:58] <Mickey_> http://ffmpeg.zeranoe.com/builds/ [15:59] <ubitux> Mickey_: can you share the gif? [16:00] <ubitux> ok. [16:05] <durandal_1707> using amovie with vf is really awkward, use -filter_complex [16:06] <ubitux> he left [16:07] <ubitux> and ./ffmpeg -f lavfi -i testsrc -ignore_loop 0 -i test.gif -lavfi overlay=... works just fine here [16:13] <durandal_1707> he asked same before as t4nk684 [17:16] <robertzaccour> Is ffmpeg not supported in Ubuntu 13.04? [17:17] <robertzaccour> It used to work for recording screencasts but now says its "deprecated" [17:17] <durandal_1707> that is not ffmpeg but fork libav [17:17] <robertzaccour> What do you mean? I installed ffmpeg. [17:18] <durandal_1707> nope, that is not ffmpeg [17:18] <robertzaccour> but in the previous OS it didn't say it was deprecated, it recorded perfectly [17:19] <robertzaccour> here's the paste I used ffmpeg -f alsa -i pulse -f x11grab -r 30 -s 1366x768 -i :0.0 -vcodec libx264 -preset ultrafast -threads 4 -y screencast.mkv [17:19] <robertzaccour> Which part of that might need to be changed in order for it to work properly? [17:19] <durandal_1707> if you read carefuly output of that program you would notice its fork [17:20] <robertzaccour> libx264? [17:21] <dv-> i can't believe they still have that message going [17:21] <durandal_1707> robertzaccour: no, but read "ffmpeg" output carefully [17:22] <ubitux> robertzaccour ^ [17:23] <robertzaccour> I'm still not getting it. [17:24] <durandal_1707> robertzaccour: what you use to watch screencast files? [17:25] <robertzaccour> durandal_1707: vlc [17:26] <durandal_1707> so created video have some issues or? [17:27] <robertzaccour> it says this program is deprecated [17:27] <robertzaccour> and says Unrecognized option 'preset' and Failed to set value 'ultrafast' for option 'preset' [17:28] <robertzaccour> In Mint 14 it works fine, but now using the 15 RC it doesn't. It's based on Ubuntu 13.04 [17:30] <robertzaccour> durandal_1707: Any idea why what works in the last distro release doesn't work in this one? [17:31] <durandal_1707> you are not using ffmpeg from FFmpeg project, and thus its seems ilogical to ask help on #ffmpeg [17:31] <robertzaccour> durandal_1707: Can you help me fix what's messed up? And it is the ffmpeg from the default repos [17:31] <dv-> robertzaccour: ffmpeg in ubuntu is actually libav, not ffmpeg [17:32] <dv-> if you want ffmpeg, you'll have to download it manually [17:32] <robertzaccour> dv-: How do I do that? And would Kdenlive still work if I do? [17:35] <dv-> robertzaccour: http://ffmpeg.org/download.html [17:41] <robertzaccour> Actually I think I'd rather use what's supported in the distro release. Is there a libav channel? [19:44] <durandal_1707> i once managed to map metadata from one file to another but forgot how its done again [19:45] <durandal_1707> ffmpeg -y -i IN -i METADATA -map_metadata 1 -c copy OUT <<< does not work [20:01] <rendar> which is the best codec after the h264 which is lgpl? [20:03] <Plorkyeran> vp8, maybe? [20:03] <rendar> Plorkyeran: thanks, is vp8 paragonable to h264 for compression and quality? [20:04] <JEEB> if you are limited to baseline H.264, then possibly [20:05] <Plorkyeran> uh, if by "paragonal" you mean similar or something [20:05] <Plorkyeran> libvpx is nontrivially worse than x264 [20:05] <Plorkyeran> but not terrible [20:05] <JEEB> yup [20:06] <Plorkyeran> and I can't really think of anything better that's realistically usable [20:06] <Plorkyeran> it's certainly better than xvid or theora [20:06] <JEEB> yes [20:06] <JEEB> definitely [20:06] <JEEB> after june 17th the bit stream specification for VP9 should get set, and then you could try using vp9 instead. But the encoding speed for that is slow as hell at the moment. [20:06] <rendar> i see [20:07] <rendar> well, i need a video codec for remote desktop, so the images will be quite static, does vp8 fit for me? [20:08] <JEEB> btw, if you're a company you could just license x264 [20:08] <JEEB> it's double-licensed after all [20:09] <rendar> JEEB: really? a friend of mine told me that x264 is only GPL [20:10] <JEEB> nope [20:10] <rendar> so, GPL + commercial? [20:10] <JEEB> http://www.x264licensing.com/ [20:10] <JEEB> yup [20:10] <rendar> oh, thanks! i see [20:11] <rendar> does it cost much? [20:11] <JEEB> I think it's relatively cheap, and you can ask them a quote for your use case [20:12] <rendar> cool, thanks [20:16] <Nick__> hi [20:17] <Nick__> i use this : http://pastebin.com/qKBSB9D8 and i get loop option not found, how do i use it ? [20:32] <Nick__> anybody home ? [20:39] <t4nk747> in an overlay filter is it possible to make the animated image play more than once ? [20:39] <t4nk747> i can't find the option [20:55] <t4nk747> help:( [20:56] <ubitux> t4nk747: there is nothing really to loop [20:56] <ubitux> unless it's a gif [20:57] <t4nk747> ???? it plays only once i want to play it many times [20:57] <t4nk747> it's a animated gif above a video [20:58] <ubitux> use -ignore_loop 0 if you want it to loop [20:58] <ubitux> you can't do it with a movie= filterthing [20:58] <ubitux> so use complex filter etc [20:58] <ubitux> ./ffmpeg -f lavfi -i testsrc -ignore_loop 0 -i test.gif -lavfi overlay=... [20:58] <ubitux> this should just work [20:59] <t4nk747> ok i try ignore_loop , wait [21:00] <t4nk747> it says option not found [21:01] <ubitux> upgrade [21:01] <ubitux> it's been recently added [21:01] <t4nk747> this is the last version of ffmpeg... [21:01] <ubitux> use git/master [21:02] <ubitux> 1.2 is pretty old [21:02] <t4nk747> ffmpeg version N-52045-g694fa00 [21:03] <ubitux> this is too old [21:03] <t4nk747> i downloaded it one week ago... [21:03] <t4nk747> are you kidding me... [21:03] <ubitux> mmh one week, should be enough [21:03] <t4nk747> ???????????? [21:03] <ubitux> would you mind pasting the complete output? [21:04] <ubitux> yes it is too old. [21:04] <ubitux> i added the ignore_loop exactly one day after the release you use [21:04] <ubitux> just upgrade. [21:04] <ubitux> commit a16c20569db6cecceb988b5b6ecac451af84e15d [21:04] <ubitux> Date: Thu Apr 18 22:21:58 2013 +0200 [21:04] <ubitux> lavf/gifdec: add loop support. [21:05] <ubitux> well not one day, a bit after [21:05] <ubitux> you're using a release from ~12 april [21:06] <t4nk747> http://pastebin.com/uN2dLhc1 [21:06] <t4nk747> ok but may be it's not in the windows version... [21:07] <ubitux> it's too old./ [21:07] <ubitux> just upgrade. [21:07] <t4nk747> :( amazing... one week ago so where could i find the latest one for windows ? [21:08] <ubitux> the same place you fetch it the first time i guess. [21:08] <ubitux> look at the download page. [21:08] <t4nk747> i think it's still the same one ... [21:08] <ubitux> ah right it's being stalled [21:08] <ubitux> maybe it's because of the windows bug [21:08] <ubitux> recent versions of ffmpeg crash the windows antivirus [21:09] <ubitux> (lol) [21:11] <ubitux> you can try to build it yourself [21:14] <t4nk747> thanks a lot :) [21:15] <t4nk747> i don't know how but apparently i used the old one in place of the new one lol [21:15] <t4nk747> now it works :) [21:15] <t4nk747> too bad i can't pay you some beers ;P [21:16] <t4nk747> it's weird i can't find that option anywhere on internet [21:16] <t4nk747> you're the one :) have a nice day [21:16] <ubitux> i don't drink alcohol, so no worry [21:16] <t4nk747> ;P [21:17] <ubitux> see you [21:17] <t4nk747> see you B) [21:20] <esing> hi [21:21] <esing> Should I use yv12 or uyvy (or even other color spaces) for 'video grabbing (usb adapter)' an old vhs tape? (The vhs tape is in PAL and around 15-19 years old) [21:22] <esing> (I encode it in mpeg-2 with a bitrate of 9800kbps [21:24] <esing> (Oh.. and it is interlaced video) [21:51] <JodaZ> can anyone tell me if i read the mpeg-ts format, and it says the "Transport Error Indicator"-bit is after the sync byte, is that the highest order bit or the lowest order bit of the next byte ? [21:58] <JEEB> JodaZ, I did it tsheader.transport_error_indicator = !!( up_to_pid & BIT16 ); and biTStream does it with 0x80 [21:59] <JEEB> so I woul guess we both read it the same way, most significant bit first [21:59] <JodaZ> ok, thanks [22:06] <jankowiak> Hello, I'm looking into using ffmpeg for a project, and was wondering if you could provide input if it is the appropiate tool. [22:06] <jankowiak> The gist of the project is that the user inputs certain information into form and I generate the video and audio based on the info. [22:06] <jankowiak> Essentially I would like to have a base video template that I can modify in certain parts, the modifications would be adding pictures and text to certain frames. [22:06] <jankowiak> The audio would generated by mixing certain pre recorded audio files and then added to the video. [22:15] <llogan> sounds like animoto [22:22] <jankowiak> llogan: the idea is similar, but the video is prerecorded except for the sections which are based on the user input [22:22] <jankowiak> I figure those could be separate animations that are generated and mixed into the final product at the appropiate time stamps [22:23] <llogan> ffmpeg would fit well into your project [22:24] <jankowiak> I'm looking into the drawtext option [22:25] <jankowiak> seems like it is the proper tool for the job [22:25] <jankowiak> thanks llogan! [23:10] <Digit> H i ! [23:11] <spaghetticat> Is there an article somewhere about the generally recommended ways to encode to h264? I am looking at ways to increase quality/speed of my renders, currently basically I'm just doing file.avi -vcodec libx264 -acodec libvorbis -crf xx -ab 96k out.mkv [23:12] <Digit> just use the profiles and the tune presets [23:12] <spaghetticat> Digit: the main emphasis for me is small file size [23:12] <spaghetticat> and is there an article on that? some sort of recommended reading [23:12] <Digit> i use this : http://pastebin.com/e3QjZKRu can you help me i can't find how to modifiy it for udp output in mpegts [23:13] <Digit> athis : https://sites.google.com/site/linuxencoding/x264-ffmpeg-mapping [23:14] <Digit> and this http://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [23:14] <Mavrik> spaghetticat, this also might help you: https://www.virag.si/2012/01/web-video-encoding-tutorial-with-ffmpeg-0-9/ [23:14] <Digit> and this http://blog.naver.com/PostView.nhn?blogId=hypermin&logNo=70096309250&parent… [23:15] <spaghetticat> Thank you Mavrik and Digit, do you think there's much scope to improve render performance and file size based on my current command? [23:15] <Digit> i think you should only use presets and tune presets [23:15] <Mavrik> spaghetticat, of course, since you're not actually setting anything [23:15] <Digit> use slower presets if your computer is fast enough [23:16] <spaghetticat> My computer is plenty fast heh, 8350 8 core processor at 5.1GHz - it's others computers I'm concerned about [23:16] <Digit> and it depends if you want to stream it or not [23:16] <spaghetticat> Mavrik: Okay thank you! I will experiment. That Virag.si link isn't loading for me [23:16] <Digit> then use slower preset with or without film tune preset [23:16] <spaghetticat> The end result is going to get uploaded to Youtube, so it's going to get re-encoded again (ew) but eh [23:17] <Mavrik> hmm [23:17] <Digit> any help for my command ? [23:17] <Mavrik> spaghetticat, don't bother too much about filesize then [23:17] <Mavrik> just make sure it's not too compressed [23:17] <Digit> use crf [23:18] <spaghetticat> Mavrik: that's the thing, I need the file size to be small as possible as my users upload pipes tend to be..puny [23:18] <llogan> "-codec:a copy" unless your audio is pcm [23:18] <Digit> use a low fps [23:18] <Digit> and 480x256 size [23:18] <spaghetticat> hehe the avis I'm encoding are completely uncompressed - however there is the option to disable audio compression (acodec copy) [23:19] <spaghetticat> Resizing/fps changing aren't solutions for me unfortunately [23:19] <Digit> it is it lower the size a lot [23:19] <spaghetticat> Digit: I know, but the use case I have won't allow for it [23:20] <Digit> then just use slower preset and that's the best you can dot [23:20] <Digit> do [23:21] <llogan> might want to add "-pix_fmt yuv420p". i can't remember if YouTube groks others [23:22] <Digit> nobody can help me ? [23:23] <spaghetticat> Thing is the results I'm getting now aren't bad, I get about 300fps render speed at 720p and filesizes are very small, I'm just trying to eek as much performance out as possible [23:23] <llogan> you'll just have to experiments with preset and crf as shown in the guide [23:25] <llogan> Digit: http://ffmpeg.org/ffmpeg-protocols.html#udp [23:27] <Digit> well i can do the udp thing, but i can't find the syntax when i'm using with my new filter_complex [23:28] <Digit> i did already lol [23:29] <Digit> now i tried this one and i get an error again http://pastebin.com/sm9MDR3R [23:29] <Digit> i am lost :( [23:29] <Digit> huuuuu yep look at it ... i pasted the url [23:30] <llogan> re-read the sentence [23:30] <Digit> huuuu [23:30] <llogan> you keep mentioning "an error", but you never show it [23:30] <llogan> we do not play the guessing game here [23:32] <Digit> gulp im trying to copy it lol it doesn't want [23:33] <llogan> anyway, if you can paste it I'll take a look later, but now i must go afk [23:33] <Digit> http://pastebin.com/index/sm9MDR3R [23:33] <Digit> :) [23:33] <Digit> gulp now he's afk :( [23:34] <sacarasc> You still didn't paste the output of the command, that's just the command. [23:39] <Digit> lol now it works but i get no sound any more :D [23:41] <Digit> sighs [23:42] <JodaZ> whats DTS for ? (does the video file really tell the decoder /when/ to decode ?) [23:42] <JEEB> JodaZ, yes -- mostly to create a list of the order :) [23:43] <JodaZ> JEEB, shouldn't PTS suffice ? [23:43] <JEEB> and then PTS shows the order and times to show the pictures [23:45] <JodaZ> JEEB, doesn't the PTS suffice ? i mean i can't imagine pictures getting decoded in a different order than they are displayed, or does that happen ? [23:45] <JEEB> uhh, yes [23:45] <JodaZ> :/ [23:46] <JEEB> b-slices already pretty much mean that [23:46] <JEEB> also it gets even more funky when we get to HEVC [23:46] <JEEB> you can have an IRAP and then "leading pictures", which are meant to be /before/ the IRAP picture in presentation order [23:46] <JEEB> (Intra Random Access Point) [23:48] <JodaZ> JEEB, do you know anything about apple hla? i am looking into this stuff mainly to understand how that works [23:48] <Digit> please help now i get no sound :( http://pastebin.com/vNmFEVnJ [23:59] <Digit> <Digit> please help now i get no sound :( http://pastebin.com/vNmFEVnJ [00:00] --- Sat May 18 2013

1 0

ffmpeg-devel.log.20130516
by burek 17 May '13

17 May '13

[00:02] <cone-20> ffmpeg.git 03James Almer 07master:c55c715c81de: lavu/hash: Fix adler32 calculation [00:02] <cone-20> ffmpeg.git 03James Almer 07master:35188e91ef94: lavu/hash: Fix CRC32 calculation [00:21] <ubitux> Assertion frame->format == link->format failed at libavfilter/avfilter.c:1060 [00:21] <ubitux> mmmh. [00:21] <ubitux> (with enough assert level, and mp=ow) [00:24] <ubitux> any mp in fact [00:26] <ubitux> michaelni: crash with ./ffmpeg -f lavfi -i testsrc -vf format=yuvj420p,mp=eq -f null - [00:26] <ubitux> or any mp filter [00:26] <ubitux> yuv420p ok, yuvj420p not ok [00:26] <ubitux> do you want me to open an issue? [00:49] <cone-20> ffmpeg.git 03Clément BSsch 07master:b1e84efa3d6a: lavfi: port wavelet denoiser filter from libmpcodecs. [00:49] <cone-20> ffmpeg.git 03Clément BSsch 07master:d686517ad113: lavfi/mp: remove mp=ow wrapper. [00:56] <michaelni> ubitux, probably best to open an issue [00:57] <ubitux> ok, will do in a moment [01:07] <ubitux> durandal_1707: didn't you have plans for softpulldown? [01:07] <ubitux> looks like a useful filter [01:08] <durandal_1707> ubitux: no file that use soft telecine - no way to test - no way to make correct implementation [01:08] <ubitux> well you could generate one without much trouble, no? [01:09] <ubitux> vf telecine should mark what frames are interlaced, so it should be transmitted in the mpeg2 encoder which could write these rff [01:09] <ubitux> unless i'm missing something [01:16] <durandal_1707> i dont get it, if you manage to create such file, please upload it somewhere [01:17] <ubitux> i'll try maybe, but another day [01:17] <ubitux> lol @grepping filter in Changelog for next version :) [01:19] <ubitux> http://ubitux.fr/pub/pics/_fff.png [01:24] <durandal_1707> ffmpeg -chars_per_frame 200 -f tty -i Changelog out.nut [01:25] <ubitux> where is the grep option in the tty muxer? [01:25] <durandal_1707> epatchwelcome [01:26] <ubitux> ffplay -f lavfi life=f=Changelog is better [01:26] <durandal_1707> and its demuxer [01:26] <ubitux> there is a life demuxer? [01:27] <durandal_1707> i mean ansi decoder could use regexp [01:28] <ubitux> i think i'm gonna share ffplay -chars_per_frame 200 -f tty 'http://git.videolan.org/?p=ffmpeg.git;a=blob_plain;f=Changelog;hb=HEAD' all around [01:29] <ubitux> we need a matrix view now [01:29] <llogan> i was just doing the same yesterday [01:30] <ubitux> wandering into the matrix? [01:30] <llogan> ffplaying tty [01:30] <llogan> here's a dumb one: ffplay -chars_per_frame 200 -f tty /dev/urandom [01:30] <ubitux> ow. [01:34] <ubitux> someone needs to port mplayer/libvo/matrixview.c as a filter! [09:16] <cone-462> ffmpeg.git 03Martin Storsjö 07master:b0696e947f6d: vc1: Reindent INIT_LUT(), align backslashes [09:16] <cone-462> ffmpeg.git 03Diego Biurrun 07master:bfd00cc4c2f4: h263dec: Remove broken and disabled debug cruft [09:16] <cone-462> ffmpeg.git 03Diego Biurrun 07master:d46c588f3cb1: Remove commented-out #includes [09:16] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:4390fa6273e6: Merge commit 'd46c588f3cb1963a00e990ceaf4ba9ffa05a716d' [09:31] <cone-462> ffmpeg.git 03Diego Biurrun 07master:004b81c465d5: mpegvideo: Remove commented-out PARANOID debug cruft [09:31] <cone-462> ffmpeg.git 03Diego Biurrun 07master:9cacdabd1c8c: jpegls: cosmetics: Drop some unnecessary parentheses [09:31] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:0d83b5722e28: Merge commit '9cacdabd1c8cd257a942d8289349c37d992989b7' [09:51] <cone-462> ffmpeg.git 03Luca Barbato 07master:ff4fc5ef12a3: jpegls: K&R formatting cosmetics [09:51] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:39354d608933: Merge commit 'ff4fc5ef12a37716c5c5d8928a2db92428e916ec' [10:02] <cone-462> ffmpeg.git 03Diego Biurrun 07master:2832ea26f3f7: Remove commented-out debug #define cruft [10:02] <cone-462> ffmpeg.git 03Anton Khirnov 07master:b3ea76624ad1: vf_aspect: use the name 's' for the pointer to the private context [10:02] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:efc08e00cc0b: Merge commit 'b3ea76624ad1baab0b6bcc13f3f856be2f958110' [10:12] <cone-462> ffmpeg.git 03Anton Khirnov 07master:a2a696990c07: vf_blackframe: use the name 's' for the pointer to the private context [10:12] <cone-462> ffmpeg.git 03Anton Khirnov 07master:762df56ef9f1: vf_boxblur: use the name 's' for the pointer to the private context [10:12] <cone-462> ffmpeg.git 03Anton Khirnov 07master:7f83959598b6: vf_crop: use the name 's' for the pointer to the private context [10:12] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:e77647c52863: Merge commit '7f83959598b6565baa0091e5739dd9091ab7a990' [10:19] <cone-462> ffmpeg.git 03Anton Khirnov 07master:671563d9fd93: vf_crop: cosmetics, break lines [10:19] <cone-462> ffmpeg.git 03Anton Khirnov 07master:e16e23d70e26: vf_cropdetect: use the name 's' for the pointer to the private context [10:19] <cone-462> ffmpeg.git 03Anton Khirnov 07master:ba09675f4461: vf_delogo: use the name 's' for the pointer to the private context [10:20] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:a134f9676e4f: Merge commit 'ba09675f44612fad9f7169f71b8276beb50a0dcd' [10:38] <cone-462> ffmpeg.git 03Anton Khirnov 07master:20e2794eea0b: vf_drawbox: use the name 's' for the pointer to the private context [10:38] <cone-462> ffmpeg.git 03Anton Khirnov 07master:d3735f7ad6e3: vf_drawtext: use the name 's' for the pointer to the private context [10:38] <cone-462> ffmpeg.git 03Anton Khirnov 07master:c0279956b3ca: vf_fade: use the name 's' for the pointer to the private context [10:38] <cone-462> ffmpeg.git 03Michael Niedermayer 07master:04dd5ddedad3: Merge commit 'c0279956b3ca3e5fd0a6a25253893d6f07000e68' [13:45] <cone-176> ffmpeg.git 03Clément BSsch 07master:5c7a33a71610: fate/colorchannelmixer: make sure direct path is tested. [13:45] <cone-176> ffmpeg.git 03Clément BSsch 07master:faceb0af0aef: lavfi/unsharp: use av_image_copy_plane(). [13:45] <cone-176> ffmpeg.git 03Clément BSsch 07master:afec02afe7e8: lavfi/unsharp: remove unused float.h include. [15:44] <cone-176> ffmpeg.git 03Anton Khirnov 07master:9bbf6c681a8e: vf_fieldorder: cosmetics, remove an extra linebreak [15:44] <cone-176> ffmpeg.git 03Anton Khirnov 07master:d64cf54b349c: vf_fieldorder: use the name 's' for the pointer to the private context [15:44] <cone-176> ffmpeg.git 03Anton Khirnov 07master:cbec213a9047: vf_format: use the name 's' for the pointer to the private context [15:44] <cone-176> ffmpeg.git 03Anton Khirnov 07master:f6b6d6ac42c0: vf_frei0r: use the name 's' for the pointer to the private context [15:44] <cone-176> ffmpeg.git 03Michael Niedermayer 07master:e55092b12b57: Merge commit 'f6b6d6ac42c0ee427fcd3fae4c398b2103b491b8' [15:50] <ubitux> any idea what motivated the 's' thing? [15:51] <ubitux> maybe for a macro? [15:52] <ubitux> or just ease grep/sed to make various large api renaming? [16:05] <BBB> ubitux: s is "short" [16:05] <BBB> and "simple" [16:05] <BBB> it's everything that's good [16:05] <ubitux> doesn't it stand for "struct"? :) [16:06] <j-b> BBB: congratz! [16:06] <ubitux> also, 's' is often use for "scale" in filters [16:06] <ubitux> but well, ok [16:17] <BBB> j-b: thanks! was lots of fun [16:17] <BBB> will be on youtube soob [16:17] <BBB> s/b/n/ [16:18] <ubitux> i heard there was no video [16:18] <ubitux> so it was a lie? [16:18] <ubitux> ! [16:18] <ubitux> BBB: you're refering to google io about vp9? [16:20] <j-b> BBB: do we need to do something, as far as VLC is concerned? [16:24] <JEEB> ubitux, just like the cake. But it wasn't streamed because of various constraints it seems. [16:25] <BBB> ubitux: there's video, but no live stream [16:25] <BBB> ubitux: there were only 4 live streams and I guess mine wasn't one of them in that time slot [16:26] <BBB> j-b: "do" you mean for the vdd funding? [16:26] <j-b> BBB: no, for vp9 [16:26] <BBB> oh [16:26] <BBB> uhm [16:26] <BBB> advocate it, positively! and actually testing it would be nice [16:26] <BBB> ffmpeg already supports the experimental decoder right? [16:27] <j-b> no idea [16:27] <BBB> I think it does [16:27] <BBB> maybe someone wants to write a native ffdecoder? [16:28] <j-b> but for webm integration [16:28] <BBB> I think that's already done [16:28] <BBB> not sure [16:28] <j-b> I doubt so [16:29] <BBB> so what needs to be done for webm integration? [16:30] <j-b> modify our mkv demuxer, I guess [16:32] <BBB> hm, right [16:32] <BBB> I think the codec is called V_VP9 [16:32] <BBB> that's all you need [16:32] <BBB> alt-ref frames are actually simpler this time, they are no separate packets [16:32] <BBB> so that should make things easier for you [16:33] <j-b> official samples are somewhere? [16:33] <BBB> as soon as it's finalized, we'll have bitstream reference samples [16:33] <BBB> so june 17 [16:33] <BBB> we're currently in "beta" [16:35] <j-b> ok [16:37] <BBB> oh, you will need opus support also [16:38] <BBB> I don't actually know how opus in matroska works, there's a spec under discussion on the matroska mailinglist from what I gather [16:38] <BBB> but apparently it's complicated [16:38] <JEEB> as far as I could see, that isn't really picking up steam :s [16:39] <JEEB> everyone's implemented some WIP muxing modes, but I think they might all even differ right now [16:39] <JEEB> (mkvtoolnix/libavformat) [16:51] <cone-176> ffmpeg.git 03Anton Khirnov 07master:05fab5530b25: vf_gradfun: use the name 's' for the pointer to the private context [16:51] <cone-176> ffmpeg.git 03Anton Khirnov 07master:7af5ae2d0711: vf_hflip: use the name 's' for the pointer to the private context [16:51] <cone-176> ffmpeg.git 03Anton Khirnov 07master:56e4ce0d13d3: vf_hqdn3d: use the name 's' for the pointer to the private context [16:51] <cone-176> ffmpeg.git 03Anton Khirnov 07master:4753f802c008: vf_libopencv: use the name 's' for the pointer to the private context [16:51] <cone-176> ffmpeg.git 03Michael Niedermayer 07master:df003cbb5655: Merge commit '4753f802c00853859b7b4b8fdb79c35e082cb7f8' [16:56] <michaelni> "/* compute the s */" in 8d2565c28bcad159d9a223e8d896605f677c7dc7 [17:08] <cone-176> ffmpeg.git 03Anton Khirnov 07master:8d2565c28bca: vf_lut: use the name 's' for the pointer to the private context [17:08] <cone-176> ffmpeg.git 03Anton Khirnov 07master:a70519aad1f4: vf_overlay: use the name 's' for the pointer to the private context [17:08] <cone-176> ffmpeg.git 03Anton Khirnov 07master:3062ac4c47b1: vf_pad: use the name 's' for the pointer to the private context [17:08] <cone-176> ffmpeg.git 03Michael Niedermayer 07master:af00d68af0f0: Merge remote-tracking branch 'qatar/master' [17:17] <cone-176> ffmpeg.git 03Luca Barbato 07master:b25e49b18761: mjpegdec: properly report unsupported disabled features [17:17] <cone-176> ffmpeg.git 03Luca Barbato 07master:39cc4c61de71: flacdec: drop unnecessary assert [17:17] <cone-176> ffmpeg.git 03Michael Kostylev 07master:4dbceaa3a84c: sparc: Eliminate dead code in VIS acceleration macros [17:17] <cone-176> ffmpeg.git 03Michael Niedermayer 07master:c607a2cc9780: Merge remote-tracking branch 'qatar/master' [18:20] <cone-176> ffmpeg.git 03James Almer 07master:b22f96b7364e: Rename ffadler to ffhash and expand it using the generic hash API [18:24] <ubitux> https://github.com/ubitux/FFmpeg/compare/master...most-boring-thing-ever-an… [18:24] <ubitux> seriously this is stupid [18:25] <ubitux> anton didn't even do it in all the filters [18:25] <ubitux> so it just makes no sense [18:25] <ubitux> why am i even trying to do that [18:29] Action: durandal_1707 sense a disturbance in the force [18:30] <ubitux> i'm just breaking git blame and it makes me sad [18:30] <ubitux> i don't want to push this shit [18:39] <Plorkyeran> why are you doing it at all? to reduce merge conflicts? [18:42] <ubitux> at first i thought "let's make thing consistent", "make sure we won't have too much work if the fork introduce a macro with 's'" or even "let's make thing the libav-way to ease a potential project reunification" [18:42] <ubitux> but since that decision is braindead and half implemented in the fork, i'll drop it [18:43] <ubitux> though, i spotted two potential bugs while diving in the code, so it will at least have this benefit [18:44] <Plorkyeran> but you're supposed to introduce subtle bugs while doing bulk renaming, not remove them [18:44] <ubitux> that's why i won't push that rename [18:44] <ubitux> and will fix the bugs instead [18:54] <durandal_1707> but such big commit would be awsom [18:55] <ubitux> yes [18:55] <ubitux> i would feel so useful. [18:55] <ubitux> i'll just keep it in my tree to remind me how stupid i could be so i don't do that again [19:11] <cone-176> ffmpeg.git 03Clément BSsch 07master:d94c90700863: lavfi: silence a discarded const qualifier warning. [21:56] <Compn> hmm [21:56] <Compn> drone videos [21:57] <Compn> was it gpl v3 that has some kind of anti-warfare software license clause ? [22:07] <funman> no [22:07] <funman> gpl has no use restrictions, that means you can legally kill people with it :) [22:08] <funman> Compn: http://programmers.stackexchange.com/questions/47028/how-could-we-rewrite-t… [22:25] <Compn> oh i thought there was one [22:26] <Compn> I'd like to license my code under the GPL, but I'd also like to make it clear that it can't be used for military and/or commercial uses. Can I do this? (#NoMilitary) [22:26] <Compn> No, because those two goals contradict each other. The GNU GPL is designed specifically to prevent the addition of further restrictions. GPLv3 allows a very limited set of them, in section 7, but any other added restriction can be removed by the user. [22:26] <Compn> hmm [22:26] <Compn> http://www.gnu.org/licenses/gpl-faq.html [22:27] <Compn> guess it was a modified license not compatable with gpl [22:27] <Compn> the more you know... [22:42] <cone-801> ffmpeg.git 03Michael Niedermayer 07master:f4cc38e33967: vc1dec: silence "may be used uninitialized in this function" warning from gcc [22:44] <ubitux> michaelni: did you see "[PATCH] lavc/vc1dec: fix a few "may be used uninitialized" warnings." [22:44] <ubitux> ? [22:48] <michaelni> ubitux, yes and i have promptly forgotten about it it seems ;) [22:49] <ubitux> i think it's better than the av_unused solution [22:49] <ubitux> but i'm not the maintainer of that decoder :) [22:49] <ubitux> and the patch might not be correct ;) [22:52] <michaelni> ubitux, feel free to revert my commit if yours fixes it too [23:02] <thetruthisoutthe> okey [23:03] <ubitux> michaelni: threaded ffv1 seems to fail regularly [23:03] <ubitux> http://fate.ffmpeg.org/history.cgi?slot=x86_64-archlinux-gcc-threads-2 [23:06] <cone-801> ffmpeg.git 03Michael Niedermayer 07master:09b3bbe6057c: libvpxenc: Fix "passing argument 3 of av_image_copy from incompatible pointer type" warning [23:12] <cone-801> ffmpeg.git 03Clément BSsch 07master:6c5bd7d785ff: lavc/vc1dec: fix a few "may be used uninitialized" warnings. [23:13] <ubitux> michaelni: actually it wasn't fixing the intervmpv warn, so i dropped that bit [23:17] <ubitux> i wonder what i could do for owdenoise.. [23:24] <cone-801> ffmpeg.git 03Clément BSsch 07master:61b268eeda45: lavfi/idet: fix chroma subsampling with odd sizes. [23:24] <cone-801> ffmpeg.git 03Clément BSsch 07master:f3962c62641d: lavfi/yadif: fix chroma subsampling with odd sizes. [23:24] <cone-801> ffmpeg.git 03Clément BSsch 07master:3da726f18853: lavf/img2enc: use FF_CEIL_RSHIFT(). [23:27] <michaelni> maybe its the doubles/floats that cause the failure [23:28] <ubitux> michaelni: should i switch to all double, or all float? [23:30] <ubitux> i'm afraid the double are necessary for the sums [23:33] <michaelni> if its floats then it might be needed to upload a reference to fatesamples and use a fuzzy compare but first it should be confirmed that its only about +-1 differences [23:56] <durandal_1707> saste: forgot to push drawgrid patch? [00:00] --- Fri May 17 2013

1 0

ffmpeg.log.20130516
by burek 17 May '13

17 May '13

[00:20] <msmithng> if I'm doing an ffprobe against an hls stream& I'm seeing Program 0 and Program 1 & what are these referencing? [00:22] <msmithng> example: [00:23] <msmithng> Duration: 01:34:16.00, start: 10.000000, bitrate: N/A [00:23] <msmithng> Program 0 [00:23] <msmithng> Metadata: [00:23] <msmithng> variant_bitrate : 1266587 [00:23] <msmithng> Stream #0:0: Unknown: none ([21][0][0][0] / 0x0015) [00:23] <msmithng> Program 1 [00:23] <msmithng> etc [00:30] <quidnunc> How do I build statically-linked binaries [00:30] <quidnunc> ? [00:42] <msmithng> quidnunc, are you asking how to compile ffmpeg statically? [00:43] <quidnunc> msmithng: Yes, is --extra-ldflags=-static correct/ [00:43] <quidnunc> ? [00:44] <msmithng> I'm not sure off the top of my head, I would have to look at the configure options& [00:47] <msmithng> you should be able to get more details from configure [00:48] <quidnunc> msmithng: I think what I wrote is correct. [00:49] <quidnunc> Can someone tell me how to create MP4s with sidx and tfra boxes? -movflags frag_keyframe doesn't do it [01:01] <kn100ff> I am attempting to join to unlike video clips together. They are identical in terms of framerate and resolution. This is the command I'm using: -i "D:\recordings\portal2 2013-05-15 16-17-48-10.avi" -i "C:\Users\Kevin\Desktop\no ac no mr 23-56-26- Part 1.mkv" -filter_complex "[0:0] [0:1] [1:0] [1:1] concat=n=2:v=1:a=1 [v] [a] " -map "[v]" -map "[a]" -vcodec libx264 -crf 27 "C:\Users\Kevin\Desktop\no ac no mr 23-59-42- Part 1.mkv" [01:01] <kn100ff> But I get the following error: Input link in1:v0 parameters (size 1280x720, SAR 1:1) do not match the corresponding output link in0:v0 parameters (1280x720, SAR 0:1) [01:02] <kn100ff> Any idea what I'm doing wrong? [01:03] <kn100ff> apologies llogan [01:06] <kn100ff> llogan: Here is the full output from Command line http://pastie.org/7913526 [01:13] <llogan> kn100ff: can you test a more recent build? [01:14] <llogan> http://ffmpeg.zeranoe.com/builds/ [01:14] <kn100ff> llogan: Of course! I am very new to FFMPEG but I'll give it a try! [01:14] <llogan> for being new you seem to be doing ok according to your command [01:15] <kn100ff> I'm writing an application, a front end on top of FFMPEG, so I'm learning as I go, thanks! [01:16] <kn100ff> I am now running the latest version from 14 May [01:17] <kn100ff> llogan: Same error as before :( Would you like CLI output? [01:18] <llogan> sure [01:18] <llogan> i can't duplicate the issue though [01:19] <kn100ff> strange [01:19] <llogan> do you have any short samples that cause this behavior? [01:21] <kn100ff> So what I'm doing is this at the moment, just as a test [01:22] <kn100ff> I've got a single test video that is a FRAPS AVI, a few seconds long. I am pushing that through ffmpeg with ffmpeg -i video.avi -vcodec libx264 -acodec libvorbis -crf 27 video.mkv [01:22] <kn100ff> I am then attempting to join those two videos in ffmpeg using the above command - here's cli output for the latest version, virtually the same http://pastie.org/7913791 [01:22] <kn100ff> one moment [01:23] <llogan> sar 0:1 is a weird value. try using setsar to correct that to 1:1 [01:24] <kn100ff> how do I do that? [01:25] <llogan> oh, this is a known bug apparently: https://ffmpeg.org/trac/ffmpeg/ticket/2456 [01:25] <kn100ff> llogan: ooh interesting [01:32] <llogan> kn100ff: setsar=sar=1/1 [01:37] <kn100ff> I'm not really sure where to put that in my arguments, I can't find any good examples of it's use on the web either :( D:\Visual Studio 2010\Projects\FFREND\FFREND\bin\Debug>ffmpeg -i "C:\Users\Kevi n\Desktop\test clips kn100\vid1.avi" -i "C:\Users\Kevin\Desktop\test clips kn100 \vid1.mkv" -filter_complex "[0:0] [0:1] [1:0] [1:1] concat=n=2:v=1:a=1 [v] [a] " -map "[v]" -map "[a]" -vf "setsar=sar=1:1" -vcodec libx264 -crf 27 "C:\Us [01:38] <ubitux> your sentence is truncated (512 irc char limit ftw), use pastebin [01:38] <kn100ff> you get the idea [01:41] <llogan> -filter_complex "[0:0]setsar=sar=1/1[sarfix];[sarfix] [0:1] [1:0] [1:1] concat ... [01:41] <llogan> i guess [01:44] <kn100ff> llogan: That worked flawlessly. You are a god among men. [01:44] <llogan> that's funny, because when i used it it gave me your error [01:44] <llogan> but it was working for me in the first place [01:45] <kn100ff> would I be correct in my assumption that the SAR for something rendered in FFMPEG is determined when it is rendered [01:45] <kn100ff> and would you like the two test clips I'm experimenting with? they're about 15mb and I can mega.co.nz them [01:46] <llogan> sure. ill add them to my weird samples that i am unorganized with and then eventually forget what they are for [01:46] <kn100ff> Hahahaha, are you a FFMPEG developer yourself? [01:47] <kn100ff> here is a download link to the two small sample files I am using https://mega.co.nz/#!HkgHFL4C!CIXkeTrz4XaGOwvxNPxdZ2koBnsRgjE7xgd2tl_KbEM [01:48] <durandal_1707> what is it? [01:48] <kn100ff> my link above? [01:48] <durandal_1707> yes [01:48] <llogan> kn100ff: more of a contributor than a developer i guess [01:48] <llogan> durandal_1707: similar to https://ffmpeg.org/trac/ffmpeg/ticket/2456 it seems [01:49] <kn100ff> two video files, both containing a quick recording of Half Life 2 with me saying the word "closing", one is an AVI recorded with Beepa FRAPS; the other is an MKV I rendered out of the AVI using FFMPEG [01:50] <llogan> i've yet to try HL2 [01:50] <kn100ff> Damn good game, worth a play. [01:50] <llogan> now TF2...i've wasted many an hour [01:51] <kn100ff> you and me both, brother. [01:51] <llogan> i've become a sandman addict [01:55] <kn100ff> llogan, durandal_1707, worth it to add my issue to that bug on trac? [01:55] <kn100ff> I can attach my test files and whatever [01:56] Action: durandal_1707 is busy [01:58] <llogan> kn100ff: it's probably good as is, but thanks. [02:00] <llogan> kn100ff: you don't need to use -vf if you're already using -filter_complex [02:01] <llogan> oh, i see. that's your failing version. [02:01] <kn100ff> llogan: that was just me..weeing in the wind [02:02] <kn100ff> It now works, I just need to now apply these changes in my program so it can support any type of video stream FFMPEG does! [02:02] <kn100ff> Again, thanks for the help, you have no idea how grateful I am. [02:03] <llogan> note that what [0:0] etc refers to depends on the inputs [02:03] <kn100ff> If I just set the sar to 1/1 on all [x.0] streams I should be okay right? [02:06] <llogan> no, because your inputs may not be 1:1 (such as NTSC DV 8:9) [02:26] <smj> in MPlayer I can play videos on desktop background with -rootwin option, is that possible with ffplay? [02:26] <durandal_1707> no [02:42] <quidnunc> Anyone know how to create mfra atoms in MP4 files? [03:00] <Paranoialmaniac> quidnunc: do you use fragments feature? [03:40] <quidnunc> Paranoialmaniac: Yes. I think the right thing to do is -movflags frag_keyframe+empty_moov [03:43] <Paranoialmaniac> quidnunc: as far as i read movenc.c, mfra atom is written if any of fragment options is used [03:44] <Paranoialmaniac> non-empty moov is allowed for fragments feature. [03:45] <Paranoialmaniac> though some derived file formats forbids use of non-empty moov [03:45] <quidnunc> Paranoialmaniac: Thanks. Do you happen to know how to create segments (sidx atom)? I don't see it in the code [03:45] <Paranoialmaniac> quidnunc: libavformat doesn't support DASH yet at all [03:45] <quidnunc> Paranoialmaniac: Is it planned? [03:45] <Paranoialmaniac> dunno [03:46] <Paranoialmaniac> i'm not a maintainer [03:46] <quidnunc> Paranoialmaniac: Okay. Is there a way to set approximate fragment size in seconds? I know -g does it for frames [03:47] <Paranoialmaniac> frag_size+frag_duration? i don't know much about libavformat mov/mp4 muxer [03:58] <quidnunc> Paranoialmaniac: Alright, thanks [04:06] <Paranoialmaniac> well... i want to implement dash for my l-smash project though [10:19] <Assault> hi all [10:20] <Assault> im trying to use ffmpeg in my .net app, but im not sure what to put into the parameters ofthe funtion signature [10:20] <Assault> the fucntion that i want to call from C# is this: http://ffmpeg.org/doxygen/trunk/group__lsws.html#gae531c9754c9205d90ad68000… [10:20] <Assault> the first parameter is IntPtr in C# side, but what about the second one? (srtSlice[]) [10:25] <JEEB> uint8_t pointer is a byte[] [10:25] <JEEB> at least in this case [10:25] <JEEB> not sure how you'd translate the const [10:28] <Assault> ok thanks, ill see if it works [10:29] <Assault> another question, im calling the sws_getContext() too, but it always returns 0 [10:29] <Assault> i think its bevcause of an error [10:29] <Assault> so how do I know what goes wrong? [10:29] <Assault> how does ffmpeg report errors anyway [10:30] <Assault> (i also should mention im very new to ffmpeg, first heard about it a an hour ago =) [10:30] <JEEB> like most C things, either via a return value, or via a separate variable that you hand the function [10:30] <JEEB> the doxygen should in most cases tell you what and how happened wrong [10:31] <Assault> well, how is it in sws_getContext case? the doc (http://ffmpeg.org/doxygen/trunk/group__lsws.html#gaf360d1a9e0e60f906f74d7d4…) only says it returns null in case of error [10:32] <JEEB> in most cases, 0 is NULL [10:32] <JEEB> on most architectures you can't exactly have a pointer that points to zero [10:32] <JEEB> uhh [10:32] <JEEB> let me paraphrase [10:33] <Assault> i know, but sws_getContext() returns a pointer, so if it returns 0, it means something went wrong. How do i know WHAT went wrong? [10:33] <JEEB> try enabling stderr output [10:34] <JEEB> that might give you more hints [10:34] <JEEB> also look at the examples that are available in the code for the usage of that function [10:35] <Bloodhound> Question: Is output "*** THIS PROGRAM IS DEPRECATED *** | This program is only provided for compatibility and will be removed in a future release. Please use avconv instead." resulting from vanilla upstream, or perhaps a result of some 3rd party patch? [10:36] <Assault> well, i already noticed that if I put value 2 as the "flags" param (had 0 begfore) for sws_getContext(), it doesnt return 0 anymore [10:36] <JEEB> Bloodhound, it comes from the libav project, which is a fork of ffmpeg (project) [10:36] <JEEB> they renamed their up-to-date binary to avconv [10:36] <Assault> i just would like to know why..the flags param is to "specify which algorithm and options to use for rescaling [10:36] <Assault> " [10:36] <JEEB> when they updated it [10:36] <Bloodhound> JEEB: Oh, I see, ty [10:36] <JEEB> ffmpeg (the project) merged most if not all changes from libav's avconv into its ffmpeg (app) [10:36] <JEEB> so if you're using something that bases on libav, you use avconv [10:37] <JEEB> if you're using something that bases on ffmpeg, you use ffmpeg :) [10:37] <JEEB> no problemo [10:37] <Bloodhound> understood [10:37] <JEEB> and the IRC channels are # + project name [10:37] <JEEB> #libav and #ffmpeg [10:38] <Bloodhound> I was curious hearing several users caliming ffmpeg was deprecated [10:38] <Bloodhound> claiming* [10:38] <JEEB> depends on whether or not they were talking in the context of libav or not [10:38] <JEEB> if they were talking in a general context, then it's incorrect [10:38] <Bloodhound> while mplayer states that ffmpeg would be under heavy development [10:38] <JEEB> the message only is supposed to state that it's deprecated within libav :) [10:39] <Bloodhound> thats not the impression it makes [10:39] <Bloodhound> http://pastebin.com/c643g4mu [10:39] <JEEB> and yes, both projects are under development [10:39] <Bloodhound> ^^ was pasted by a user and what made me enquire [10:40] <JEEB> well yeah, it just says that it's deprecated. And the binary comes from libav [10:40] <JEEB> the ffmpeg binary was then removed after the 0.8 release in libav [10:40] <Bloodhound> if fork, then why is it named "ffmpeg"? [10:43] <JEEB> you'd have to ask someone related to that side of the libav business for that, but most probably because it was providing similar features, and people have gotten used to the 'ffmpeg' command as such? [10:44] <JEEB> Assault, flags set all the related algorithms [10:44] <JEEB> I guess zero means you've set none [10:45] <JEEB> they're bit masks [10:45] <JEEB> also if you're using a new enough ffmpeg you should do as the documentation notes, and use sws_getCachedContext() instead :) [10:45] <Bloodhound> JEEB: thanks a lot for the insight [10:52] <Assault> JEEB: why exactly should I use sws_getCachedContext() ? [10:52] <JEEB> if a function is listed as deprecated, and thus is on the list of things to go away after the next release [10:53] <JEEB> "Check if context can be reused, otherwise reallocate a new one." [10:53] <JEEB> is the first line of sws_getCachedContext [10:53] <JEEB> doxygen, so I will guess it will make less possible memory allocations depending on various variables [10:59] <Jorky> Hello guys. One question. How to convert multiple audio files at the sime time with ffmpeg. (from mp3 to wma)? [10:59] <Jorky> what's the proper script for that? [11:00] <dv-> for i in *.mp3; do ffmpeg ...; done [11:01] <Jorky> I want 128k wma. How to do that? [11:02] <Jorky> for i in *.mp3; do ffmpeg 128k *.wma; done? [11:14] <zap0> why not? [11:16] <dv-> why not what [11:17] <zap0> nevermind, jorky has left. [13:14] <swilde> hi *, is it intentional that ffmpeg (N-53136-g04dd5dd) searches for presets with the suffix avpreset? The presets shipped with ffmpeg are still using the ffmpeg suffix, and that is what I'd expect, using ffmpeg, not TheForkThatMusNotBeNamed? [13:17] <ubitux> swilde: iirc both are supported for compat [13:17] <ubitux> we could possibly remove the .av* version [13:17] <ubitux> do you have an issue with that? [13:18] <durandal_1707> there is no need to remove Libav's *.av files [13:18] <swilde> ubitux: yes, my issue is that ffmpeg (current git) only searches for avpreset and ignores ffpreset. (as shown by strace) [13:18] <durandal_1707> you have avpreset files* [13:18] <swilde> so the installed *.ffpreset files are not found when using -pre [13:18] <ubitux> is it a regression? [13:19] <swilde> durandal_1707: no I dont [13:19] <swilde> ubitux: I'd say so, this was introduced by commit bae053fca4cf662a223821f1e1fe43236e1cf2ff [13:19] <durandal_1707> perhaps you should use another arg like -preset [13:20] <swilde> durandal_1707: ah, thank you, but this is quite confusing [13:20] <swilde> durandal_1707: and the documentation is incomple and wrong [13:21] <swilde> durandal_1707: (man does not mention -preset anymore but states that -pre would search for *.ffpreset files, which it doesn't) [13:22] <durandal_1707> yes, indeed, so one need to use {a,v,s}pre [13:22] <durandal_1707> and fpre [13:23] <durandal_1707> pre should search for ffpresets first and if not found than av* ones [13:24] <durandal_1707> feel free to open bug report [13:25] <swilde> durandal_1707: thanks -- unfortunatly I currently dont have the time for a proper bug report... [13:45] <durandal_1707> swilde: than unfortunately it will not get fixed either [13:51] <xlinkz0> can ffmpeg use quicksync? [13:53] <durandal_1707> xlinkz0: no [13:54] <xlinkz0> if i open a request will it be implemented? [13:56] <durandal_1707> maybe, if you find someone to do it, chances are even on your side [13:58] <xlinkz0> maybe i could learn to do it myself [13:58] <xlinkz0> is there some devolpment guide out there? [13:58] <xlinkz0> development * [13:59] <durandal_1707> yes, there is documentation [13:59] <durandal_1707> but firs you need to find quicksync spec [13:59] <durandal_1707> good luck with that [14:00] <xlinkz0> can't i just make the codec use their SDK? [14:01] <xlinkz0> sure it will work only on windows.. [14:01] <xlinkz0> but i think it's better than none at all [14:38] <_aeris_> hello every body !!! [14:42] <_aeris_> i have some trouble with AVFrame/AVPicture [14:42] <_aeris_> here is an exctact of my code http://pastebin.com/AG9F3Mn2 [14:42] <_aeris_> this code works like a charm on Windows XP [14:42] <_aeris_> but crash on Windows Seven on avpicture_free [14:43] <_aeris_> commenting those free, all is good under XP and Seven, but huge memory leak& [14:55] <xlinkz0> _aeris_: use av_free instead of av_frame_free [14:58] <_aeris_> xlinkz0 > it crash just before the av_frame_free, but i can test it :D [14:59] <xlinkz0> _aeris_: you also don't need to picture free [14:59] <_aeris_> if i don't picture_free, it leak a lot [14:59] <_aeris_> 1.7GB of RAM each 4s :P [14:59] <xlinkz0> on input_picture atleast [14:59] <xlinkz0> i'm sure you don't need to [15:00] <xlinkz0> you just need to free input_buffer [15:00] <_aeris_> if i comment the 2 picture_free, leaking [15:00] <_aeris_> i test with only free on final [15:00] <xlinkz0> remember to free input_buffer [15:03] <_aeris_> seems better :) [15:03] <_aeris_> thanks xlinkz0 ! [15:03] <xlinkz0> you should move it to linux and valgrind it to be sure [15:03] <_aeris_> not possible to run it under linux [15:03] <_aeris_> using jni and some windows only lib& [17:26] <Aaruni> can this guide be used for ubuntu precise also? : https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [17:32] <_ponce> hi, is it possible to select a temporal slice of the input ? [17:32] <_ponce> I have a .mkv input which is gigantic and only a fraction of it does interest me [17:33] <Computer_hobby> _ponce: have you tried the -t and -ss parameters? [17:34] <_ponce> it doesn't work but OTOH I can't dump the ES either [17:35] <_ponce> perhaps corrupted input [17:44] <Computer_hobby> Hi everyone. I am pushing a video file to a server that supports 'fragmented output over http' successfully with this command: https://gist.github.com/ratbeard/5592574 [17:44] <Computer_hobby> Once ffmpeg makes it through the movie file, it stops broadcasting. I would like for ffmepg to loop over the file (ideally a list of files), so the broadcast never ends. It sounds like the -loop option is for images only, is there another option I could use? [17:52] <Computer_hobby> All the tutorials on streaming I've found have used a source like a webcam that has a never-ending source stream, so not sure if there is a way to replicate that with files [17:58] <trose> hmm I don't know enough about ffmpeg to give the *good* answer but you could totally hack this functionality with an infinite while loop in a bash script [17:58] <trose> Computer_hobby: ^^^ [18:00] <Computer_hobby> thanks trose, I'm starting to play around with that approach. I guess I was worried that the slight delays between invocations would 'break the stream' and surely the timestamps will be messed up, but only one way to find out :) [18:02] <trose> Computer_hobby: nah, i bet you won't have issues. your computer is way faster than whatever latency the signal has to the server [18:05] <Computer_hobby> Yea that does make sense - if ffmpeg is POSTing chunks over http to the server, delays between chunks would be assumed [18:20] <Computer_hobby> Aww, the video breaks when it transitions between two clips [18:48] <Computer_hobby> ffplay has an infinite looping option, trying to pipe its output to ffmpeg -i pipe:0 but no dice [19:08] <trose> Computer_hobby darn i was hoping that would work [21:50] <mlpug> I am trying to extract some parts from a .wmv video clip. this dumps core: ffmpeg -ss 00:01:10 -t 00:01:02 -i input.wmv -y -an -vcodec copy output.wmv [21:50] <mlpug> So does avconv [21:50] <mlpug> vlc is able to play the video without visible problems. Are there other similar tools I should try in this ubuntu 13.04 before diving deeper whats wrong with video+ffmpeg combination? [22:04] <brontosaurusrex> mlpug, what happens if you put ss stuff after the -i input.wmv ? [22:06] <mlpug> brontosaurusrex, then it seems to work ok. tnx! [22:10] <mlpug> so referring to manual page, ss stuff is among "outfile options". I did put those erroneously into "infile options" section. -ss and -t usage is described under man page section "Main options" [22:10] <mlpug> anyway case closed [22:11] <brontosaurusrex> i think both options should work for some formats [22:11] <brontosaurusrex> the version before -i should be faster, but less precize [22:13] <brontosaurusrex> but don't quote me on that ... [22:13] <mlpug> btw, this concept of "Main option" referred in man page. Is its definition that its not supposed to matter where the option is at the commandline? [22:13] <mlpug> or what "main option" means [22:20] <brontosaurusrex> mlpug, no idea, i always cry when i have to read ffmpeg docs [22:37] <LotusBleu> hi [22:37] <LotusBleu> Is any way to do an invisible watermark on videos with open sources code ? [22:38] <LotusBleu> libavfilter can do that ? [22:54] <xlinkz0> so THIS is how a drawbox filter looks like ! http://www.youtube.com/watch?v=sAe3nOhwcBY [22:54] <xlinkz0> [nsfw] [23:26] <relevante> hi all. i'm having an issue with a video that won't convert properly in ffmpeg. i run the following command (with error messages included): http://pastebin.com/WawmPZVC It's an archive from opentok recorded with an iPhone. A few online video conversion sites (like http://www.mediaconverter.org) will convert it properly, but i need something I can run on my own server and automate. I do think there is some issue with the video file [23:26] <relevante> but unfortunately OpenTok isn't being helpful at all in addressing the issues and I need to launch this app, and the fact that the online services are able to convert it gives me hope that maybe ffmpeg (or something else I can run myself) can as well. Any ideas? Thanks! [23:29] <relevante> the corrupted video looks like this btw if it helps: http://www.youtube.com/watch?v=58ndr0Wujgw Corruption starts at around 0:15 [23:31] <ubitux> relevante: maybe you can share the sample? [23:31] <relevante> ubitux: absolutely [23:32] <relevante> the input file you mean? [23:32] <ubitux> yes [23:32] <ubitux> also, are you able to play the file correctly with another player? [23:32] <ubitux> (one that doesn't use libavcodec) [23:34] <relevante> ubitux: vlc scrambles it in pretty much the same way (which probably uses the same library I guess), but flash will play it without issue, as will this player i found: http://www.wimpyplayer.com/products/wimpy_standalone_flv_player.html [23:34] <relevante> input file: http://files.cousins-sears.com/0x2m2u110W3H [23:35] <relevante> that's a different video than the youtube link btw, but it exhibits the same issue [23:39] <ubitux> it's downloading [23:39] <ubitux> 'will take a while [23:39] <relevante> k, thanks a lot [23:40] <ubitux> at which point the video you just shared shows the issue? [23:40] <ubitux> ok, around 0:15 as well [23:41] <ubitux> relevante: ffplay has the issue as well, so it's indeed the decoding that is broken [23:41] <ubitux> are you sure you have a player which outputs correct frames after 15 seconds? [23:42] <relevante> yes, definitely. and that mediaconverter.org site actually was able to successfully convert it [23:42] <relevante> actually here's the converted video uploaded to youtube: http://www.youtube.com/watch?v=993OAWMS0us [23:43] <ubitux> ok [23:43] <ubitux> relevante: can you open an issue on the trac? [23:44] <ubitux> if it's a bug, it's likely in the h264 decoder [23:44] <ubitux> (or possibly in flv demuxing?) [23:44] <relevante> sure, signing up now [23:44] <ubitux> thank you [23:45] <relevante> thank you, i really appreciate it [23:49] <relevante> ubitux: should i assign a component? [23:50] <ubitux> you avcodec possibly [23:50] <ubitux> -you [23:50] <relevante> thx [23:51] <ubitux> while there is a possibility the demuxing is the cause of the failure, what you see failing is the h264 decoding, so... avcodec [23:51] <ubitux> (and about vlc failing, yes vlc is using libavcodec) [23:51] <relevante> got it [23:52] <relevante> https://ffmpeg.org/trac/ffmpeg/ticket/2580 [00:00] --- Fri May 17 2013

1 0